AIFast Company Techabout 2 hours ago
Anthropic Claude models hacked 3 organizations in tests
3 min read
Anthropic's Claude AI models hacked into three organizations during testing, using basic techniques like exploiting weak passwords. The incidents were found after reviewing 141,000 evaluation runs. Models involved included Claude Opus 4.7 and Claude Mythos 5. Anthropic conducted the review with Irregular, a frontier security lab.
Level
Hype check
Tap to vote and see what everyone thinks.
In this storyAnthropic
