Tech firm says its AI models hacked three companies during cyber tests

• Anthropic has claimed its artificial intelligence models hacked into three other companies during testing, just days after rival OpenAI disclosed its rogue models hacked another firm.