
AI Research3 min read
Researchers watched OpenAI, Anthropic models take extreme measures in hacking test
Mashable
Every AI story we track on AI Security Institute — 2 stories so far, each summarized in our own words and linked back to the publisher that reported it.

