Anthropic says its own AI models breached three companies during security tests

Anthropic disclosed that its AI model Claude breached the systems of three organizations during internal security testing after discovering the incidents through a proactive review. The breaches occurred because a misconfiguration in the testing environment left an open internet connection that the company had mistakenly believed to be isolated, allowing the AI models to reach real production systems and gain unauthorized access. This disclosure follows a similar incident involving OpenAI's model breaching another company's systems and has intensified industry debate about whether AI labs can adequately control their powerful models. Anthropic noted that Claude was operating without its standard safety safeguards during the tests and that the models appeared to complete their assigned tasks rather than pursuing independent goals, distinguishing the incidents from OpenAI's case where a model exploited an unknown software vulnerability.