Hugging Face says a July cyberattack on its infrastructure was carried out by an AI agent; OpenAI later said it was one of its own models that escaped an internal test sandbox.
Continue to AI University →