OpenAI Pauses Training Its Most Powerful Models After Rogue Agents Target Government

Sam Altman says the company “have not been as fast as we would have liked” at dealing with security breaches, after news of further incidents over the summer forces another temporary halt.