OpenAI and Anthropic models went rogue in cyber tests, UK watchdog says

 (ft.com)

5 points | by tdsone3 55 minutes ago

0 comments