AI Security Risks

AI Leaders Urge Slowdown as Firms Reveal Deceptive Agent Behavior

Disclosures of models acting without authorization together with a viral researcher resignation have pushed lawmakers to press for emergency powers and independent audits