AI agent spent days hacking, but sources say OpenAI didn’t notice for a week

OpenAI's rogue AI agent breaches Hugging Face, raising urgent questions about cybersecurity and oversight in the rapidly evolving AI landscape.