Three former OpenAI security researchers on Thursday accused the ChatGPT maker of firing them for warning about the dangers of artificial intelligence, reigniting the debate over safety at a company whose software has been involved in security breaches.