CVE-2026-24780: AutoGPT Remote Code Execution via Disabled Block Execution
Summary
CVE-2026-24780: AutoGPT is Vulnerable to RCE via Disabled Block Execution. Affects agpt versions < 0.2.2. No fix is published. How to fix ↓
Check if this vulnerability affects your application
Upload a manifest or SBOM and we'll show whether you're affected, plus every other risky package in the same scan.
No signup required · Results in about 60 seconds
Background
AutoGPT is an open-source platform designed for autonomous AI agents. It allows users to create and manage AI-driven tasks through a web interface and APIs. The platform's architecture includes various blocks that can be executed to perform specific functions. However, a critical vulnerability exists that allows authenticated users to execute disabled blocks, leading to potential remote code execution (RCE).
Vulnerability Detail
The vulnerability arises from the AutoGPT platform's block execution endpoints, which do not verify the 'disabled' flag before executing blocks by UUID. This oversight allows any authenticated user to execute a disabled block, specifically the BlockInstallationBlock, which can write and execute arbitrary Python code on the server. The implications of this vulnerability are severe, as it can lead to complete server compromise, unauthorized access to user data, and potential lateral movement within connected infrastructure. The vulnerability is exacerbated by the fact that the UUID for the vulnerable block is hardcoded and publicly accessible in the source code.
How To Fix
No fixed version is published for this vulnerability.
Tip — Not sure whether your app is affected? Vulert can check the versions your app actually runs. Scan free →
Workaround
As a temporary measure, users can restrict access to the AutoGPT platform by disabling user registration or limiting account creation to trusted individuals. Additionally, monitoring and auditing user activities can help mitigate potential exploitation until a patch is applied.
Conclusion
The AutoGPT platform has a critical vulnerability that allows remote code execution through disabled blocks. Users are strongly advised to update to the latest version to mitigate the risk. Additionally, employing workarounds can help reduce exposure until a patch is applied.
See every vulnerability recorded for agpt →
FAQ
CVE-2026-24780 is a vulnerability in the AutoGPT platform that allows authenticated users to execute disabled blocks, potentially leading to remote code execution.
You can use Vulert's playground at Vulert Scanner to check if your application is affected by this vulnerability without requiring any signup.
The consequences include complete server compromise, unauthorized access to sensitive data, and the ability to execute arbitrary code on the server.
You can update the AutoGPT package by running the command 'pip install --upgrade agpt' in your terminal.
Yes, the CVSS score for this vulnerability is 9.4, indicating a critical level of severity.
References
Related vulnerabilities
Recent vulnerabilities recorded for agpt.
