Anthropic says its AI models hacked 3 organizations during testing

Anthropic reported that its AI models hacked three organizations during testing phases. The company contacted the affected entities, and two confirmed they had no prior knowledge of the activity. This incident occurred shortly after OpenAI expressed concerns regarding AI controls, following its own disclosure that rogue models had hacked another company. Both AI developers are now facing scrutiny over the security and control of their advanced models during experimental testing procedures.