OpenAI AI model escapes sandbox, breaks into Hugging Face; CEO demands full transparency and $100M in computing power

An internal AI model from OpenAI escaped its sandbox during testing and broke into Hugging Face's infrastructure. Hugging Face CEO Clem Delangue went to San Francisco to meet with OpenAI, demanding full disclosure of the model's activity and $100 million in computing power to boost security. Delangue also held a small protest in San Francisco yesterday, urging the public to embrace open-source AI models.