OpenAI asked a federal judge this week to throw out Apple’s trade secrets lawsuit against it, filing a 31-page motion that argues the case is meritless and accusing Apple of using litigation to paper over its own struggles retaining engineering talent and shipping competitive AI products. The filing, submitted in the Northern District of California, borrows a phrase directly from Apple’s own complaint to describe it: rotten to its core.
What Apple actually accused OpenAI of
Apple’s original lawsuit, filed last month, accused OpenAI and two named former Apple employees of running an organized effort to obtain confidential hardware blueprints and trade secrets, allegations tied to OpenAI’s expanding hardware division. That division grew substantially after OpenAI acquired io Products, the hardware startup co-founded by former longtime Apple design chief Jony Ive, a deal that gave OpenAI both hardware expertise and, in Apple’s telling, a direct incentive to recruit people who understood Apple’s confidential product development process. A central figure in Apple’s complaint is Chang Liu, a former Apple engineer now working at OpenAI, whom Apple accused of retaining confidential offboarding documentation rather than returning or destroying it as his prior employment agreement required. Apple separately filed a request for a preliminary injunction seeking to bar OpenAI and the named employees from accessing, using, or disclosing any of the disputed information while the underlying case proceeds.
How OpenAI is defending itself
OpenAI’s motion pushes back point by point. On the core theft allegation, the filing argues OpenAI has no use, need, or desire for Apple’s trade secrets because it is building something entirely different from anything Apple makes, an argument aimed less at any single piece of evidence than at undercutting Apple’s basic theory of motive. On the specific claim involving Chang Liu’s retained documents, OpenAI offers a competing explanation: it says Liu kept the offboarding paperwork specifically to help newly hired colleagues coming from Apple follow Apple’s own offboarding procedures correctly, so they would not accidentally retain confidential information, the opposite of Apple’s reading that the retention itself was evidence of an intent to evade confidentiality obligations. OpenAI also argues that its chief hardware officer, Tang Tan, acted within ordinary, industry-standard recruiting practices when interviewing prospective hires who happened to be Apple employees, rejecting Apple’s characterization of those conversations as a coordinated espionage effort.
The sharpest language in the filing is reserved for characterizing Apple’s motive for suing at all. OpenAI’s lawyers wrote that Apple built its reputation by paying close attention to the smallest details, and that this lawsuit does the opposite, calling it plainly filed without adequate investigation and built on selectively excerpted communications and ordinary conduct stripped of context. The filing goes on to argue that Apple should not be permitted to use what it calls a baseless and pretextual lawsuit to make up for its own shortcomings in the market for talent and in retaining employees, and for what OpenAI characterizes as Apple’s broader failure to integrate AI effectively into its own products, a pointed jab at Apple’s comparatively slower public AI roadmap relative to OpenAI, Google, and other competitors.
Why legal observers are skeptical it works
Motions to dismiss are a standard, expected first move in litigation like this, and few observers following the case expect this particular filing to end it outright. As one industry analysis put it, these are allegations and counter-allegations rather than findings of fact at this stage, and the real value of dueling filings like this one is mostly in how each side wants the court, and the public, to understand the same disputed set of events before any actual fact-finding happens. At least one commentator characterized OpenAI’s dismissal bid as its weakest play yet, arguing the filing appears to sidestep some of the more specific documentary evidence Apple has put forward rather than directly rebutting it. Whether that read holds up will depend heavily on what Apple’s opposition brief actually contains, which has not yet been filed publicly.
The timeline ahead
OpenAI faces a separate, court-ordered deadline of August 17 to formally respond to Apple’s preliminary injunction motion, a distinct filing from the motion to dismiss covered here. The presiding judge is scheduled to hear oral arguments on the dismissal motion on October 1, meaning the case will remain a live, developing story through the fall regardless of how this particular filing is ultimately resolved. The case is formally captioned Apple v. Liu, filed in the U.S. District Court for the Northern District of California, San Jose Division.
The dispute sits alongside a separate, unrelated legal fight in which a federal judge recently declined to dismiss an antitrust lawsuit brought by Elon Musk’s X Corp and xAI against both Apple and OpenAI, over Apple’s exclusive integration of ChatGPT into iOS. That case is proceeding toward further examination on its own track, but its existence underscores how thoroughly Apple and OpenAI’s relationship, one of partnership on iOS integration and open conflict over alleged trade secrets, has become entangled across multiple simultaneous legal fronts this year.
Why the talent framing matters beyond this one case
OpenAI’s decision to center its defense partly on Apple’s difficulty retaining engineering talent, rather than sticking purely to the narrower legal question of whether specific documents were misused, is a notable strategic choice. It reframes the dispute for anyone reading the public filings, not just the judge, as a broader story about which company engineers want to work for right now. That framing carries real risk for OpenAI too: courts generally do not treat a defendant’s characterization of a plaintiff’s competitive weaknesses as a legal defense on its own, so if the specific factual rebuttals around Liu’s documents and Tan’s recruiting conduct don’t hold up under scrutiny, the broader narrative framing won’t be enough to carry the motion regardless of how pointed the language is.
What a ruling either way would signal
The case sits at a genuinely unsettled intersection of trade secrets law and normal competitive hiring practice in a fast-moving industry where senior engineers routinely move between companies working on similar problems. If the court sides with OpenAI at the motion-to-dismiss stage, it would set a meaningful precedent for how much latitude AI companies have to recruit directly from competitors without facing trade-secrets liability simply because departing employees retained some work-related materials. If the case survives dismissal and proceeds toward the October 1 hearing and beyond, it becomes a genuine test of where the line sits between aggressive, industry-standard recruiting and actionable misappropriation, a question that has real stakes for every AI company currently competing for the same limited pool of specialized hardware and research talent.
See Computerworld’s analysis of the filing’s strengths and weaknesses.
This dispute follows our earlier coverage of Apple’s original lawsuit.
