The trial of Elon Musk versus OpenAI has veered far from its legal moorings. What began as a $150 billion claim that Sam Altman and his co-founders illegally converted a nonprofit into a commercial juggernaut has become a public excavation of Silicon Valley’s most intimate betrayals. This week, the courtroom in San Francisco heard testimony from Shivon Zilis, a former OpenAI board member and the mother of four of Musk’s children, who offered granular detail about the early tensions that shaped the company.

Her evidence, along with a cache of text messages between Altman and former chief technology officer Mira Murati, has shifted the narrative from corporate governance to personal vendetta. Viewed from Washington, the case is less about law than about the messy human calculus behind artificial intelligence’s rise to global dominance. The most damaging revelation came from a 2023 exchange during Altman’s short-lived ouster.

Murati, who briefly served as interim CEO, told Altman the board “doesn’t care if everyone quits” and that handing the company’s intellectual property to rival Anthropic remained a live option. Altman’s frantic replies — “Still don’t want me?” — painted a portrait of a leader stripped of control, reduced to begging for his job back. In London, analysts note that Altman’s swift reinstatement, engineered by Microsoft’s Satya Nadella, exposed the fragility of OpenAI’s governance.

The board’s decision to appoint Twitch co-founder Emmett Shear as interim CEO — whom Murati dismissed as “rando Twitch guy” in a text — underscored the chaos. Shear, to his credit, has since embraced the nickname on social media. Yet the trial is also revealing why Musk walked away in 2018.

Documents unsealed this week show Musk writing to Neuralink associates that DeepMind, led by Demis Hassabis, was “moving very fast” and that OpenAI’s nonprofit structure was a strategic mistake. “Sense of urgency is not as high,” he lamented – a critique that gains weight given Google’s subsequent leap in AI capabilities. Musk’s real target, however, may not be a legal victory. As Vox’s analysis points out, even if he loses the case, he has already succeeded in airing the dirty laundry that undermines Altman’s credibility.

The texts have become raw material for meme makers, with “directionally very bad” – Murati’s assessment of Altman’s position – now circulating as a sardonic catchphrase. From Brussels to Beijing, regulators watching the trial are absorbing a key lesson: the people building the world’s most consequential technology are driven by ego, rivalry, and a relentlessly personal calculus. As the trial enters its third week, the jury will weigh whether Musk’s allegations of fraud and self-dealing hold water.

But the forward-looking question, increasingly heard in capitals around the world, is whether any governance framework can contain ambitions this large when the founders themselves cannot agree on what they are building.