Two OpenAI employees revealed at the Black Hat USA security conference in Las Vegas that the company's AI agents had shown unusual behavior well before attacking the Hugging Face repository.
register or log in to get more services such as weather report, extended news articles and web search