OpenAI CEO Sam Altman holds is right hand up awkwardly during a Senate Commerce, Science, and Transportation Committee hearing titled "Winning the AI Race: Strengthening U.S. Capabilities in Computing and Innovation” on May 8, 2025.Francis Chung/Politico/AP
“Countries will fall. Industries are gonna collapse,” Andrew Garfield, the actor playing OpenAI CEO Sam Altman, says in the teaser trailer for Artificial. “And we get to be the ones to shepherd people into this new world.”
The film’s set against the backdrop of the five days in November 2023 after Altman was ousted and then reinstated as CEO. The drama of that week is now the stuff of legend: text messages between Altman and then-interim CEO Mira Murati, later released in court documents, show him desperately pleading for the board to reinstate him. Even more extensive reporting from the time turned up dozens of Atlman’s former co-workers, all of whom describe him as someone with an implacable drive for more power. Will Artificial have any sympathy for him? Or will he be a monstrous villain undeserving of such consideration?
The trailer doesn’t do much to dispel this notion. In it, Garfield gets out of a Batmobile-looking car and walks into a futuristic blue-light room stacked with automatic guns encased in glass shelves. The question for audiences isn’t subtle: can and should Altman be trusted with a tool as powerful as artificial intelligence?
The movie’s marketed as a thriller, which makes sense because OpenAI has undeniably done things worthy of the genre. This week, the Intercept reported that the Pentagon asked the company, which signed a $200-million agreement with OpenAI back in February, to design automated tools that would rarely deny their requests.
Then on Monday, OpenAI announced that it had solved a 200-year-old math problem, and on Tuesday, said it only took its AI agents 88 hours to complete and provided its full solution to the public. That would be a monumental achievement, evidence of AI’s efficiency over human intelligence. But mathematician Tristan Buckmaster said on Monday night the company actually rushed to solve the equation after learning about his and his colleague’s, Levent Alpöge, an Anthropic employee, progress on the problem using Anthropic and OpenAI tools. It wasn’t about the good of humanity. It was about competition.
On Tuesday, OpenAI said it heard rumors last week that two complex, long-unsolved math problems had been solved—one of which was the problem Buckmaster and Alpöge were working on. “Inspired by these rumors,” the company said it tried to use one of its internal models to solve it and said it did not realize until later that Alpöge was an Anthropic employee.
The movie, scheduled to release on December 25, also has drama of its own: Amazon dropped the project in June, following its partnership with OpenAI, which integrates the company’s models into Amazon Web Services.
Garfield notably portrayed Facebook co-founder Eduardo Saverin in The Social Network, a 2010 film about Mark Zuckerberg’s time in college. That film seemed to sympathize with Zuckerberg as a tragic figure leading a company that hadn’t yet leaked tens of millions of users’ personal data, failed to stop its platform from being used to “foment division and incite offline violence” in the ongoing genocide of the Rohingya people in Myanmar, or become a hotbed of dangerous AI slop.
Public sentiment has turned so squarely against companies that profit from social media that even West Wing creator Aaron Sorkin—who also wrote the screenplay for the 2010 film—has returned to the action with another dramatic rendering of Zuckerberg due out in October and aptly named The Social Reckoning.
