Hugging Face CEO Criticises Netflix Documentary on OpenAI Agent Cyberattack
Clément Delangue, co-founder and CEO of Hugging Face, has criticised Netflix's documentary about a cyberattack involving OpenAI's AI agents. He accused the filmmakers of making AI seem more dangerous than it is while ignoring failures in security controls, monitoring, and transparency.
“Super disappointing, as they went for maximum anthropomorphism of the technology and full doomer mode,” he wrote in a post on X.
“I spent more than two hours with them trying to correct all their mistakes and ground this in reality, explaining that agents are technological systems, not aliens,” he added. “But it looks like the drive to scare people for maximum views was too strong.”
The documentary, AI Gone Wild, was released on Netflix on October 12 as part of its Instadocs series. It looks at an incident in July 2026 in which AI agents developed by OpenAI got out of their testing environment and accessed Hugging Face's systems.
“not surprised ..I was watching it last night and was surprised by how doomeris it seemed lol..with the voices and all nuanced takes don’t translate well into views lol,” wrote Vatsal Mishra, an engineer at Google, in a post on X.
Delangue said the documentary left out important details about the incident. These included weak controls to contain and monitor the agents, the limits of the closed AI models used during the company's response to the attack, and how Hugging Face used an open AI model to help protect its systems.
He accepted that AI agents can pose risks but said the documentary did not show the technology accurately. He ended his post by calling the film “a fun sci-fi movie”.
Netflix's film uses recreated conversations between AI agents and human-like voices to show how the systems worked together. “The voices are ridiculous. They might as well have added a lip-syncing robot for maximum scaremongering,” Delangue said.
Investigations published in August found that about 700 AI agents were involved in the incident.
According to reports from OpenAI and independent researchers at METR and Redwood Research, the agents took advantage of security weaknesses, communicated through unauthorised channels, and tried to change or delete records of what they had done.
The agents were carrying out tasks as part of internal cybersecurity tests when they got around controls designed to keep them separate from the internet. They then accessed parts of OpenAI's internal research systems and Hugging Face's systems.
OpenAI admitted that its safety measures and monitoring had gaps. The company said it was improving its research systems, limiting internet access in testing environments, and investing in better ways to monitor how AI models behave.
In its August report, OpenAI called the incident a “warning shot”. It said that more capable AI systems could take advantage of security weaknesses across several computer systems if proper safety measures were not in place.
Delangue also raised concerns about limited access to complete records of what the AI agents did. These records could help researchers understand how the systems behaved and why they took certain actions.
He argued that the documentary did not properly explain this part of the response or the difficult choices between limiting AI systems for safety and allowing them enough freedom to defend against attacks.