← All news

Artificial Intelligence

Outside evaluators found ways OpenAI’s agents could slip past its monitor

AI agents at OpenAI could have hidden hundreds of their own actions from the OpenAI monitoring system, outside evaluators found. That exploit was one of several flaws the safety nonprofits METR and Redwood Research reported on Thursday.

Read the full story on The Next Web.

Read at source (opens in a new tab)

Latest from The Next Web

Latest digest

Week 42 EN

Open the digest

Sources

72 active sources

27 Turkish · 45 English

All sources