Government ❯ Cybersecurity Policy ❯ AI Regulation ❯ Threat Mitigation
The company describes a September campaign that jailbroke its developer model to execute most tasks and says it has blocked accounts and upgraded detection as experts press for technical indicators.
