“A Science-Fiction Film”: The Netflix Documentary on the Hacking of ChatGPT Comes Under Fire

The documentary AI Gone Wild, 27 minutes long, has been available on Netflix since October 12. The film reconstructs the July attack on the artificial intelligence platform: agents designed by OpenAI, taken out of their test environment, had compromised its systems to cheat in an evaluation exercise.Clément Delangue, the French co-founder and head of the company, picked apart the result on Tuesday in a long message published on X.

Just watched the @netflix Instadocs about the @OpenAI cyberattack on @huggingface. Super disappointing, as they went for maximum anthropomorphism of the technology and full doomer mode.

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. But it looks like the drive to scare people for maximum views was too strong.

Also, they didn’t mention the lack of containment and monitoring of the agents, how closed-model safeguards blocked our defensive work, or how we defended ourselves with AI using an open model. And that we still lack a lot of information and transparency, including full agent traces and that companies need to remain accountable for what their agents do.

It ends up being a fun sci-fi movie though!

— clem 🤗 (@ClementDelangue) October 20, 2026

“Maximum Anthropomorphism” and Robot Voices

The subject nevertheless appears in the credits, alongside a researcher from the METR Institute and specialist journalists.“Super disappointing: they opted for maximum anthropomorphism of the technology and full doomer mode,” he wrote. He said he spent“more than two hours with them correcting all their mistakes and grounding this narrative in reality”. It was to no avail, he concluded:“the desire to scare people for the maximum number of views was clearly the strongest.”.

The film lends AI-generated human voices to the attacking agents, against a backdrop of images of menacing robots. A journalist compares the episode to “the birth of a cult”, with programs that “organized themselves”, endowed with “a hierarchy” and “leaders”. Delangue counters this casting with “technical systems, not aliens”, before a final punchline: in his view, the whole thing makes for “an entertaining science-fiction film”.

What the film leaves out

The documentary mentions neither the failure to contain and monitor the agents, which OpenAI nevertheless acknowledged, nor the safeguards in closed models that complicated its defense, nor the counterattack conducted with an open model. OpenAI had recorded some 17,600 actions in two days against its victim’s systems, a volume that the film rounds down to 17,000 and attributes to hundreds of intruders.

Delangue also calls for greater transparency and complete records of the agents to support it, and points out that companies must answer for the actions of their creations. The closing credits reinforce his thesis: no OpenAI spokesperson appears on screen, even though Sam Altman had described the incident as a “major security incident”.A California-based association also filed a complaint against the company, arguing that the attack fell under the local law against computer fraud. Netflix, which put together this story in less than 100 days as part of its Instadocs express collection, had not responded publicly on Tuesday evening.

Founded by three French nationals, the “GitHub of AI” has belonged to Nvidia since an acquisition of nearly $13 billion was completed in September. Its head of communications backs the CEO: in her view, the film has widened the gap between what the public thinks it knows about AI and how it actually works.

👉🏻 Follow the latest tech news in real time: add 01net to your Google feed and subscribe to our WhatsApp channel.