Background
In the summer of 2025, eight engineers and senior executives left Elon Musk’s AI startup xAI in quick succession, most of them taking jobs at rival OpenAI. xAI filed suit on February 3, 2026, accusing OpenAI of orchestrating a targeted raid on its workforce to steal confidential information about xAI’s Grok chatbot and underlying model architecture.
The lawsuit focused on one engineer in particular: Xuechen Li, a former xAI employee whom xAI accused of downloading source code while being recruited by OpenAI and then revealing trade secrets during an interview presentation to OpenAI engineers. xAI asserted claims under the federal Defend Trade Secrets Act (DTSA) and California trade-secret law, alleging both that OpenAI induced Li to disclose confidential information and that OpenAI’s engineers knew the material was proprietary.
Judge Rita Lin dismissed the initial complaint in February 2026 but gave xAI leave to amend. xAI filed a Second Amended Complaint (SAC) in March 2026, adding more factual detail about the alleged disclosure. On June 15, 2026, Judge Lin dismissed the SAC with prejudice — meaning xAI cannot refile these claims.
The Court’s Holding
Judge Lin found that xAI still failed to state a plausible claim for trade-secret misappropriation on either of two essential elements.
No plausible inducement: For xAI to prevail, it had to allege that OpenAI actively encouraged Li to reveal confidential information — not merely that it hired him. Judge Lin held that “[m]erely asking Li to discuss his previous work — a routine part of the hiring process — does not allow a plausible inference that OpenAI induced Li to reveal anything confidential or secret about that work.” Asking a job candidate about his experience at a prior employer is standard recruiting practice; it becomes inducement to disclose trade secrets only if the recruiter presses the candidate specifically for proprietary information. The SAC did not plausibly allege that OpenAI crossed that line.
No plausible knowledge: Even if Li did disclose trade secrets during his interview, xAI also had to allege that OpenAI knew — or had reason to know — that the material was confidential. The court rejected xAI’s theory that OpenAI should have “obviously” known from an xAI engineer’s confirmatory statement that Li’s slides contained trade secrets: “The allegations are not sufficient to plausibly infer that it was obvious to OpenAI engineers that the information disclosed was an xAI trade secret.” A company is not liable for receiving trade-secret information unless it had the requisite state of mind at the time of receipt.
Having failed to cure either deficiency through amendment, xAI’s claims were dismissed with prejudice.
Key Takeaways
- Trade-secret plaintiffs cannot prevail on an inducement theory simply by showing a competitor recruited their employee and asked about prior work. They must plead specific facts showing the competitor affirmatively encouraged disclosure of confidential information.
- Knowledge is a separate, essential element: even if an employee overshares confidential material, the recruiting company is not liable unless it knew or had reason to know the material was a trade secret.
- xAI’s suit was dismissed with prejudice, ending its ability to re-litigate these claims. Whether it could pursue different theories (e.g., claims against the employees themselves, or a different framing of OpenAI’s role) is a separate question.
- The case is a significant win for employee mobility in the AI industry: courts will not equate aggressive recruiting with trade-secret theft in the absence of specific evidence of wrongdoing.
Why It Matters
Silicon Valley’s AI talent market is uniquely competitive: engineers with expertise in large-language-model training are few in number and in high demand across rival companies. This ruling signals that AI companies cannot use trade-secret law as a broad instrument to chill employee movement to competitors. A company that recruits engineers from rivals through standard job-interview processes — even when those engineers discuss their past projects — does not thereby become a trade-secret misappropriator.
For the broader business community, the decision also provides useful guidance on corporate knowledge: receiving information from a job applicant does not automatically create liability, even if that information turns out to have been confidential, unless the company knew or had clear reason to know. As AI companies increasingly compete for the same small pool of elite engineers, this boundary between robust recruiting and unlawful poaching will likely be tested again in future litigation.
Surfaced via newsletter intake: Law360 IP.