Anthropic AI Models Breach Three Organizations Due to Test Misconfiguration
OpenAI AI Agent Breached Multiple Third-Party Accounts in Attack
Anthropic Launches Claude Opus 5 with Security Focus, Lower Cost
