Two companies, two trade secrets claims
Apple filed suit against OpenAI in July in a case court filings refer to as Apple v. Liu, alleging trade secret theft connected to a former Apple employee who left to join OpenAI, according to reporting referenced by Bloomberg and The Verge. As part of that case, OpenAI has agreed not to use any trade secrets Apple claims were stolen while litigation continues, and to collect forensic images from the devices of 14 OpenAI employees, a significant concession that signals the court sees enough substance in Apple's claims to warrant real evidentiary access rather than dismissing the complaint early. Both companies have been actively recruiting from each other's ranks, so a dispute like this one was arguably a matter of when, not if, given how often confidential product knowledge travels in people's heads rather than in files a company can fully control.
OpenAI, in its own filings, is not simply defending. Reuters reported that OpenAI alleges a different former employee, one who left OpenAI to join Apple, sent more than 100 files about an unreleased OpenAI product to a personal account and gave a presentation about it to Apple leadership. Apple disputes the accusation, saying it found no evidence of the specific file OpenAI identified. Both halves of this dispute are now live in the same litigation.
Why a judge is taking this seriously
TechCrunch reported that a judge ordered OpenAI to turn over its hiring records as part of the case, a discovery step that suggests the court wants to understand the broader pattern of movement between the two companies, not just the specific individuals named in the complaint itself. Hiring records discovery is the kind of order courts issue when they suspect an isolated incident might actually be a pattern worth examining systematically, rather than a one-off dispute between two departing employees who happened to get caught.
The Verge's reporting on the case suggested the judge may be leaning toward Apple's side of the dispute, though the litigation is ongoing and that read should be treated as an early signal rather than a verdict anyone should rely on yet. What is already clear is that a federal court is treating the exchange of personnel between a hardware giant and a frontier AI lab as a serious vector for trade secret exposure, with real evidentiary consequences attached, rather than dismissing it as a routine talent market story that does not warrant close judicial attention.
The forensic device order is the real headline
Agreeing to collect forensic images from 14 employees' devices is a meaningfully bigger commitment than it sounds. It means OpenAI consented to a process that can expose internal communications, file access history, and personal account activity for a specific group of employees as part of court-supervised discovery. Companies do not agree to that lightly, which suggests OpenAI's legal team judged it a better outcome than fighting the request and losing anyway.
For any enterprise that has ever worried about what happens when an employee leaves for a competitor with sensitive knowledge in their head or on their laptop, this is a concrete example of how far litigation can go to find out. It also sets a practical precedent: courts are willing to order intrusive device-level discovery in AI trade secrets disputes, which raises the stakes for how companies manage offboarding and device policies today, not after a lawsuit is filed.
The talent war has legal teeth now
Frontier AI labs and Big Tech have been aggressively recruiting from each other for two years, often with compensation packages explicitly designed to pull senior researchers and engineers across company lines. This case is the clearest evidence yet that the same recruiting intensity that builds headcount also builds legal exposure, in both directions, since both Apple and OpenAI are now accusing former employees of the other of taking something they should not have.
That symmetry matters. It is not a story about one company behaving badly while another plays defense. It is a story about an industry where talent moves fast enough, and stakes are high enough, that trade secret disputes are becoming a normal cost of doing business rather than a rare event. Any company competing for the same pool of AI talent should assume the same risk applies to its own hiring and departures.
What this means for enterprise hiring and IP policy
Legal and HR teams at any company recruiting AI talent, whether from a frontier lab, a competitor, or an adjacent industry, should treat this case as a template for what can go wrong and what courts will actually order in response. That means tightening offboarding procedures, clarifying device and account ownership before an employee's last day rather than after, and making sure new hires are briefed on what they cannot bring with them, in writing, as a condition of employment.
It is also a reason to revisit how your own contracts handle AI vendor and partner relationships where personnel regularly move between organizations, including subcontractors and consultancies that touch multiple competitors in the same space. If forensic device audits and consent injunctions are now a realistic outcome of a trade secrets dispute between two of the best-resourced companies in technology, smaller enterprises should assume they have even less room, financially and operationally, to absorb a similar dispute if it happens inside their own supply chain or partner network.
