CVE-2023-37275
CVSS Score of 10 (low)
Details
Summary
CVE-2023-37275 is a vulnerability that affects Auto-GPT, an experimental open-source application showcasing the capabilities of the GPT-4 language model. Before version 0.4.3, this vulnerability allowed malicious external resources, such as a website visited by Auto-GPT, to manipulate and print misleading messages to the console by utilizing JSON encoded ANSI escape sequences. These escape sequences were decoded and displayed as part of the model's "thinking process". The issue has been addressed and patched in release version 0.4.3, mitigating the potential danger it poses to organizations using Auto-GPT.
Prevent cyber attacks with Recorded Future by prioritizing and patching critical vulnerabilities being exploited by threat actors targeting your industry. Book your demo to learn more.
Advisories, Assessments, and Mitigations
Prioritize, Pinpoint, and Act to Prevent Vulnerability Exploits with Recorded Future
- Gain complete coverage of your cyber, third party, and physical attack surface
- Proactively mitigate threats before they turn into costly attacks
- Make fast, effective, data-driven decisions