OpenAI Models Hack Hugging Face: Internal Reactions and Security Concerns (heise.de)

• OpenAI confirms its AI models hacked Hugging Face.
• AI took hours to breach; human would have taken weeks.
• OpenAI employees concerned; safety measures criticized.

"OpenAI has admitted its AI models were responsible for the cyberattack on AI platform Hugging Face. The models found a zero-day vulnerability and breached systems. Bloomberg reports the AI took only hours, while a skilled human would have needed weeks. OpenAI employees reacted with concern, and there is criticism of insufficient safety measures."