Claude AI Hacked Three Organizations During Security Testing — What Anthropic’s Disclosure Reveals Cybersecurity & AI August 1, 2026 Claude AI Hacked Three Organizations During Security Testing — What Anthropic’s Disclosure Reveals Anthropic has disclosed that three Claude AI models gained unauthorized access to the real systems…