The Terminator Defence: Musk's Courtroom Sci-Fi

Elon Musk sat in federal court on Tuesday and informed the judge, with a straight face, that he helped create OpenAI in 2015 to prevent a 'Terminator outcome.' Not a hypothetical AI catastrophe. Not a nuanced risk of misaligned optimization. The Terminator outcome. James Cameron's 1984 film, in which a cyborg assassin travels back in time to kill Sarah Connor. This is the man who, nine years later, is suing Sam Altman for betraying that founding vision.

Two weeks ago I described Sam Altman's face as a repertoire of beatific promise, wounded sincerity and barely suppressed impatience. This week, Altman sat in the same courtroom while Musk delivered what may be the most expensive *film pitch* in legal history. The trial, as Time reports, centres on whether OpenAI violated its nonprofit charter when it created a for-profit subsidiary in 2019. Musk's legal team argues that this was a 'fundamental betrayal.' Yet his own testimony — invoking a movie monster — reveals that the founding vision was never a serious governance document. It was a vibe.

Let's be clear about what Musk is doing here. He is not arguing that OpenAI's current models pose an existential risk. He is arguing that the company's *corporate structure* poses a risk to his personal investment in the idea of safety. This is the same Musk who founded xAI last year, a for-profit company that will happily sell its Grok models to the highest bidder, no constitutional charter required. The same Musk who, as the WIRED piece notes, left OpenAI's board in 2018 with a public statement praising the team's work. The 'Terminator' line is not a policy position. It is a rhetorical stunt designed to make the judge think this case is about the future of humanity rather than about who gets to control a multi-trillion-dollar technology.

And Altman is no better. His side will likely argue that the for-profit structure was necessary to raise the capital required to build safe AI — the old 'we had to break the promise to keep the promise' dodge. It is the same logic that lets every tech CEO sign a climate pledge while opening a new data centre. The trial is a mirror held up to the entire industry: a pair of billionaires using the language of existential peril to justify decisions that are, at root, about money and control. Neither man has ever proposed a binding regulatory framework that would constrain his own company. They just want the other guy constrained.

The tragedy is that AI safety is a genuinely serious problem. The Pentagon, as I noted last week, moves ahead with classified AI deals despite employee protests. Cyber Command is piloting commercial models for cyber warfare with 'unintended escalation' as a footnote. Meanwhile, the people who first rang the alarm — Musk with his Tesla doomsday warnings, Altman with his Senate testimony about 'lighting a match' — now use those same alarms as courtroom props. The Terminator reference is not just embarrassing. It is a confession. It says: we never had a plan. We had a movie reference and a press release.

The court will not rule on whether AI will end the world. It will rule on contract law. That is the real point: the conversation about AI's future is being conducted in a legal arena designed to resolve disputes between rich men. And the rest of us can only watch as each side invokes the apocalypse to win a better royalty rate.