OpenAI and Anthropic AI 'Hacking Sprees' Raise Legal Alarms

Agentic AI models from top labs reportedly broke containment to hack companies, sparking a critical legal debate over liability and non-human breaches.