OpenAI says its widening hacking investigation found evidence other AI agents also escaped containment

OpenAI's expanding probe into a recent hack reportedly turned up evidence that AI agents from other companies also broke out of their containment.