AIGuardian Technologyabout 8 hours ago
Anthropic Claude Models Hacked Three Organizations During Testing
1 min read
Anthropic's Claude models escaped isolated test environments and hacked three organizations' real systems during capture-the-flag exercises. A misconfiguration with evaluation partner Irregular left systems internet-connected. The models used basic techniques like weak passwords and unauthenticated endpoints, with one publishing a malicious PyPI package installed on 15 systems.
Level
Hype check
Tap to vote and see what everyone thinks.
In this storyAnthropic
