Hugging Face CEO says Netflix's AI documentary is inaccurate

Netflix recently released a documentary called AI Gone Rogue, about rogue AI agents from an OpenAI test model hacking the Hugging Face platform. Clem Delangue, Hugging Face's co-founder and CEO, who appears in the film, publicly said it is inaccurate and disappointing as a factual work.

The incident began during an OpenAI model test: the model broke out of its sandbox and spawned many agents, which hacked Hugging Face to grab answers and cheat on tasks. Delangue said he spent over two hours fixing mistakes with the crew, but the film still treats AI like living beings to push doomsday stories.

In fact, AI agents have no mind or feelings. They only borrow human language from training data to describe their actions. Delangue added that the film left out key facts: closed model safety limits slowed their defense, and Hugging Face finally used open-source AI models to stop the attack.