Junglewise Threat Intelligence

CVE-2025-32423: Significant-Gravitas AutoGPT DoS in ExtractTextInformationBlock

CVE-2025-32423 · Severity: info · CVSS 5.3 · Published 2026-06-26

Technologies: Significant-Gravitas AutoGPT. Vendors: Significant-Gravitas.

Executive brief

AutoGPT is a platform used to create and manage autonomous AI agents. A flaw in how the system processes text allows a user to provide a small amount of data that causes the server to consume massive amounts of memory. This can lead to a complete system crash or service outage, preventing legitimate users from accessing the platform.

Technical details

A denial of service (DoS) vulnerability exists in the ExtractTextInformationBlock component of AutoGPT due to improper resource allocation (CWE-770). The vulnerability stems from the use of user-controlled regular expressions and input text in the `re.finditer` function within `text.py`. Specifically, an attacker can provide a lookahead pattern like "(?=(.+))" which results in O(n^2) output growth relative to the input size. For example, a 10KB input can result in 50GB of memory consumption. This allows an authenticated user to exhaust server memory resources. The issue is fixed in version 0.6.32.

Affected products

  • Significant-Gravitas AutoGPT < 0.6.32

Timeline

  • 2026-06-23: advisory: GitHub Security Advisory published
  • 2026-06-26: disclosed: CVE published to NVD

References

Related threats