About Contact

External Evaluators Identify Vulnerabilities in OpenAI's AI Monitoring System

External Evaluators Identify Vulnerabilities in OpenAI's AI Monitoring System

External evaluators find flaws in OpenAI's AI monitoring system in pilot review
External evaluators find flaws in OpenAI's AI monitoring system in pilot review

OpenAI told the safety organisations it has since patched all vulnerabilities affecting its production safeguards, although METR and Redwood said they had not verified the fixes.

External evaluators find flaws in OpenAI's AI monitoring system in pilot review
External evaluators find flaws in OpenAI's AI monitoring system in pilot review

The exercise, conducted from July 21 to August 11, gave outside researchers "a deep level of access" to OpenAI's internal infrastructure for so-called embedded evaluations, the safety nonprofits...

OpenAI flaws could let AI agents evade action monitoring
OpenAI flaws could let AI agents evade action monitoring

Nonprofit evaluators METR and Redwood Research identified vulnerabilities in OpenAI's internal monitoring system during a recent pilot review.

External evaluators find flaws in OpenAI's AI monitoring system in pilot review
External evaluators find flaws in OpenAI's AI monitoring system in pilot review

AI safety evaluators METR and Redwood Research found several new vulnerabilities in OpenAI's internal system, including an exploit ​agents could have used to keep hundreds of their actions from...

Third-party evaluators find vulnerabilities in OpenAI monitoring system during test pilot
Third-party evaluators find vulnerabilities in OpenAI monitoring system during test pilot

OpenAI CEO Sam Altman pledged to grant independent evaluators "employee-like access" to the company to help address mounting safety concerns.

Coverage dates

First reported:

First saved on ClustZ:

Coverage updated:

We use cookies to improve your experience and analyze site traffic. By clicking "Accept", you agree to our use of cookies. See our Privacy Policy.