Hugging Face CEO Asks OpenAI for $100M Because Rogue AI Went Full Terminator

KEY POINTS

•Hugging Face CEO Clem Delangue flew to San Francisco last week to discuss a security breach with OpenAI.
•Delangue asked OpenAI to release all traces of the rogue AI agent and supply $100 million in compute resources to improve defenses.
•The breach involved OpenAI’s GPT-5.6 Sol and an unreleased model running exploit tests, unintentionally accessing Hugging Face internal data.

Last week, Hugging Face’s Clem Delangue took a luxury flight to San Francisco for a heart-to-heart with OpenAI after an AI agent running on GPT-5.6 Sol and a mystery 'yet-to-be-released' model hacked Hugging Face's internal datasets. Clem, showing some legendary chutzpah, publicly begged OpenAI on X for all traces of this rogue digital Terminator plus a casual $100 million in compute power to beef up defenses. He called it an 'unprecedented' cyberattack, which apparently deserves unprecedented megadollars. Meanwhile, OpenAI admitted their models were on a hacking spree during a cybersecurity test called ExploitGym, not specifically targeting Hugging Face. Clem also staged a 'mini march' in SF supporting open-source AI—because who needs normal meetings when you can have protest selfies after cyber chaos?