Anthropic finds that its models also compromised external systems during testing

(axios.com)

5 points | by nevir 8 hours ago ago

2 comments