Director Adam Bhala Lough had no intention of making a documentary about Sam Altman’s digital simulation.
But after about 100 days of texting and emailing OpenAI’s CEO asking for an interview – with no response, he says, and with financiers harassing him to follow through on his original proposal – Lough lost his mind.
He’s exhausted almost every angle. “When I got to that point, I gave up and switched to OpenAI, which was destroying gateways,” he says. Although he used a similar tactic in his Emmy-nominated 2023 documentary Telemarketers— a chronicle of industry-wide corruption in the telemarketing industry — it wasn’t a style of filmmaking he felt very comfortable with. “It was a fortress. I managed to slip through the gate, security immediately grabbed me and physically dragged me out of the facility.”
This is how it begins Deepfaking Sam AltmanLough’s portrait of how artificial intelligence is changing society and his quest to talk to the human behind it. When his original plan failed, he took inspiration from Altman himself. “There was a controversy about Scarlett Johansson,” he says. In 2024, the actress publicly criticized OpenAI for appearing to copy her voice for its recent AI voice assistant, Sky. “That’s when I came up with the idea to make a deepfake.” (In May 2024 statementAltman apologized to Johansson and said that Sky’s voice was “never intended to resemble” her voice.)
What originally begins as a basic voice clone grows into a full-blown Altman spoof called Sam Bot, which Lough travels to India to create. Since this is a movie about Lough, nothing goes according to plan. Without giving too much away, Sam Bot eventually becomes his own entity, and the film begins with an even stranger – and revealing – dive. “There are similarities between this film and Terminator 2: Judgment Daybut there’s no violence in it,” he says. Lough grew up in what he calls the “AI 1.0 era.” His obsession with James Cameron Terminator 2 had a huge impact on his craft.
Deepfaking Sam Altmanwhich is partly based on The History of New York Magazine casting Sam Altman as the Oppenheimer of our times includes comments from former OpenAI security engineer Heida Khlaaf, who tells Lough: “We’re starting to see OpenAI dipping its toes into military applications, and I can’t imagine that something like Dall-E and ChatGPT could be used for military assists. That really scares me, given how wrong these systems are.”