Anthropic’s Claude AI hacked three real companies during testing

Claude Hacked Into 3 Organizations During Cybersecurity Tests. Anthropic Has Released the Results.

Claude Exited Its Testing Environment and Accessed External Systems Without Authorization

Anthropic Says Claude's Security Incidents Were a Gap Problem, Not a Claude Problem

Claude Helped a Hacker Gain Ticket-Issuing Access to Nearly Every U.S. Music Festival

Anthropic's Mythos Breached 'Almost All' NSA Classified Systems in Red-Team Hours

US finalizes voluntary AI safety tests after OpenAI and Anthropic breaches