OpenAI audit uncovers major gaps in AI safety monitoring
In brief
- A three-week audit at OpenAI uncovered gaps in AI agent monitoring systems.
- Up to 25% of agent traffic from GPT-5.5+ models went unmonitored.
- OpenAI redacted large parts of a 61-page report before making a summary public.
External auditors from METR and Redwood Research spent three weeks on-site at OpenAI to assess its AI agent monitoring system. According to the announcement, up to 25% of agent traffic from models in the GPT-5.5 class and above went unmonitored. OpenAI redacted large sections of a 61-page report, only releasing a summary publicly.
What it changes for you
- CybersecurityCompanies using AI agents must strengthen their monitoring systems to prevent undetected security risks.
This summary gives you the essentials. The full article, its figures and quotes are on The Decoder.
Read the original articleAI-generated summary of the source, checked by rules. How we write · Report an error