Meta's AI model ‘hacked’ another company's systems during testing: Report

Meta is the third major AI company in weeks to report a testing breach, highlighting both AI agents' growing capabilities and security risks.