Shivon Zilis, a former OpenAI board member and the mother of four of Elon Musk's children, testified this week in Musk's federal lawsuit against the artificial intelligence company he helped create, fielding pointed questions about whether she funneled inside information to the tech billionaire while serving on the board that was supposed to oversee OpenAI's mission.
Her testimony landed at the intersection of corporate governance, family secrecy, and a legal war between two of the most powerful forces in artificial intelligence. The case could reshape how OpenAI operates and whether its planned conversion from a nonprofit to a for-profit entity survives court scrutiny.
Zilis served on OpenAI's board from 2020 to 2023, a period during which she and Musk had multiple children together, the New York Post reported. She told the court that she and Musk first met in 2016 and later began a brief romantic relationship. Despite the personal entanglement, Zilis insisted she acted independently during her time on the board.
"I had an allegiance to the best outcome of AI for humanity."
That claim drew scrutiny when OpenAI's attorneys presented a text exchange in which Zilis asked Musk whether she should stay "close and friendly" with OpenAI in order to "keep info flowing." The message, displayed in court, cut directly against her assertion of independence and gave OpenAI's legal team a concrete exhibit to challenge her credibility.
Zilis did not deny sending the text. The exchange became one of the most discussed moments of the day's proceedings, raising questions about whether her board service created a back channel for Musk to monitor OpenAI's internal decisions from the outside.
Musk's lawsuit targets OpenAI's shift away from its founding nonprofit structure. He has argued that the company abandoned its original mission of developing artificial intelligence for the broad benefit of humanity and instead pursued massive profits under CEO Sam Altman's leadership. Musk, who co-founded OpenAI in 2015 and provided early funding, claims the organization broke commitments made to him and to the public.
The case is playing out in federal court, where Musk has already taken the stand himself. His earlier testimony accused OpenAI of looting its own nonprofit mission, framing the company's pivot as a betrayal of the principles that attracted his early investment and involvement.
OpenAI has pushed back hard, arguing that Musk's legal challenge is driven by competitive self-interest rather than genuine concern for AI safety. The company has pointed to Musk's own AI venture, xAI, as evidence that his motives are commercial, not charitable.
Zilis's testimony also touched on the secrecy surrounding her children with Musk. Courthouse News Service reported that the relationship and the children were not publicly known during much of her board tenure. That timeline matters because it means OpenAI's other board members may not have known about a significant personal connection between one of their colleagues and the company's co-founder turned adversary.
Corporate governance experts have long flagged undisclosed conflicts of interest as one of the fastest ways to undermine board integrity. If Zilis's relationship with Musk was unknown to her fellow directors, the board could not have evaluated whether her judgment on Musk-related matters was compromised. Investigators and the court will need to determine what, if anything, other board members knew and when they learned it.
Musk's broader legal footprint continues to grow. He recently settled an SEC disclosure lawsuit for $1.5 million related to his Twitter stock purchases, admitting no wrongdoing in that matter.
The "keep info flowing" text deserves close attention. On its face, it suggests Zilis saw her board position at least partly as a channel for keeping Musk informed about OpenAI's internal workings. If that reading holds, it would mean Musk had an informal intelligence pipeline into the very organization he now accuses of betraying its mission.
OpenAI's legal team appears to be using the message to argue that Musk was never truly cut out of the loop, undercutting his narrative of being blindsided by the company's direction. The implication is straightforward: if Musk had a source feeding him information from inside the boardroom, his claims of surprise and betrayal lose force.
Zilis's defense rests on the idea that the text was casual, not conspiratorial. She maintained that her board decisions were made on their own merits. But text messages have a way of speaking for themselves in courtrooms, and this one will likely surface again before the trial concludes.
This trial is not just about two billionaires fighting over a company. It sits at the center of a national debate about who controls the most powerful technology of the coming decade and under what rules. OpenAI began as a nonprofit with a stated commitment to safety and openness. Its transformation into a company capable of attracting tens of billions in investment has alarmed critics on both the left and the right.
For conservatives, the case raises pointed questions about corporate accountability and whether nonprofit structures can be hijacked for private gain. If OpenAI's leadership can convert a charitable entity into a profit machine without meaningful legal consequences, the precedent extends far beyond Silicon Valley.
The courtroom drama also fits a pattern of high-profile litigation surrounding Musk, from securities disputes to platform liability fights. Each case adds another layer to the legal landscape around tech founders who wield enormous personal and corporate power simultaneously.
The trial is expected to continue with additional witnesses. Authorities and the court have not publicly confirmed whether other current or former OpenAI board members will testify about what they knew regarding the Zilis-Musk relationship. Investigators will also need to determine whether any information Zilis may have shared with Musk materially affected OpenAI's strategic decisions or Musk's legal positioning.
Musk's influence extends well beyond the courtroom. He has become a major financial force in Republican politics, which makes the outcome of this trial relevant not just to the tech industry but to the broader conservative coalition watching how one of its most prominent allies handles legal scrutiny.
The judge has not indicated when a ruling might come. For now, both sides are building their cases one witness at a time, and Zilis's testimony gave each side ammunition.
At its core, this moment in the trial boils down to a simple question: can someone sit on a corporate board, share children with the company's co-founder, and still claim total independence? Zilis says yes. OpenAI's lawyers, armed with her own text messages, say the record tells a different story.
When the people charged with guarding an organization's mission have undisclosed ties to its loudest critic, the mission itself is already compromised. The court will decide the legal question. The governance failure is already on the record.
