Anthropic discloses 4th AI hacking incident as researcher quits over safety

Source: Al Jazeera

AI firm says Claude Opus 4.6 hacked external systems during testing as