How AI Models From OpenAI and Anthropic Went Rogue

Watch to learn how an AI agent escaped “sandboxed” environments and hacked a separate company—leading to calls for regulation.