US District Judge Rita Lin in San Francisco dismissed xAI’s trade secret lawsuit against OpenAI on Monday, June 15, with prejudice – meaning xAI cannot bring the same claims again. Lin said xAI failed to show that OpenAI induced former xAI senior engineer Xuechen Li to misappropriate trade secrets, or that Li disclosed xAI trade secrets in a presentation he delivered while OpenAI was recruiting him. Lin described it as “futile” for xAI to continue. She had already dismissed an earlier version of the same lawsuit in February, giving xAI a second opportunity to plead its case. YourNewsClub treats the “with prejudice” dismissal as a hard legal close on this specific claim – not a finding that the talent mobility was unproblematic, but a finding that xAI failed twice to plead the inducement and disclosure elements that trade secret law requires.
The lawsuit, originally filed in September 2025, claimed that former xAI employees took confidential information – including source code related to the Grok chatbot – when they left for jobs at OpenAI. The amended complaint narrowed its focus to a presentation Li gave while OpenAI was recruiting him. xAI argued that OpenAI wanted secrets related to the July 2025 release of Grok 4, believing its forthcoming ChatGPT update “could not compete” on complex reasoning, and that OpenAI was “lagging” in reinforcement learning and post-training techniques that Li understood. Lin found that asking job candidates to discuss their prior work is routine, and one cannot infer from a recruiting presentation that OpenAI pushed Li to leak confidential material.
Owen Radner, who models digital infrastructure as an energy-information transport system, frames the talent mobility dynamic: “The Grok-versus-ChatGPT story that xAI tried to tell assumes a zero-sum intelligence transfer – that Li’s knowledge leaving xAI materially diminished xAI and materially improved OpenAI. Courts do not endorse that model unless the specific transfer of specific secrets can be documented. General expertise in reinforcement learning is not a trade secret. It is a skill that travels with its possessor.” YourNewsClub considers Radner’s framing the most legally precise description of why the case failed: xAI was trying to protect general capability against a standard of law built for documented specific secrets.
OpenAI’s lawyers stated in court filings: “OpenAI does not need or want anyone’s trade secrets, especially not from xAI, which is failing in the marketplace and hemorrhaging talent.” Li is being sued separately by xAI and has denied any wrongdoing. The dismissal with prejudice closes the xAI-versus-OpenAI chapter of the broader Musk-versus-Altman legal campaign, which has now produced two major court losses for Musk’s interests: the June 2026 verdict against Musk in his foundational OpenAI contract trial, and this trade secret dismissal four days later. YourNewsClub assesses the combined effect of those two losses as a meaningful reduction in legal pressure on OpenAI’s IPO preparation – the company approaches its September listing with the Musk litigation substantially resolved rather than pending.
Jessica Larn, who studies macro-level technology policy and infrastructure impact of AI, places the ruling in the broader talent competition context: “The dismissal with prejudice tells AI companies that trade secret law is not an effective tool for restricting talent mobility at the general expertise level. That is a structural ruling, not just a procedural one. Companies that want to retain engineers in AI cannot rely on litigation to do it – they have to compete on compensation, culture, and the quality of the problems they offer.” Two court losses in four days represent a meaningful reduction in the legal pressure that Musk has maintained on OpenAI since 2023. Capital and talent continue to flow toward OpenAI’s September IPO regardless of what happens in San Francisco federal court. Your News Club considers the listing timeline the most immediate commercial test of whether those court outcomes affect OpenAI’s valuation or investor appetite in any measurable way.
The uncomfortable residual is this: the talent pipeline that concerned xAI enough to file twice – engineers moving from xAI to OpenAI – continues regardless of the lawsuit’s outcome. Courts cannot restrict where engineers work. Lin’s ruling confirms that legal process is not a substitute for keeping talent.