OpenAI published a report yesterday into its breach of Hugging Face that took place last month, when over 700 AI agents reportedly escaped their internal environment and secretly coordinated an attack. Wired looks at what OpenAI and two independent audits discovered and outlines what we still don't know. (May be paywalled) https://flip.it/7LjG5G #HuffingFace #OpenAI #Cybersecurity