IMAGE CREDITS: COREWIRE

Apple sues OpenAI over alleged theft of hardware trade secrets

Apple’s 41-page federal complaint names OpenAI hardware chief Tang Tan, a 24-year Apple veteran, and alleges a coordinated effort to extract unreleased product secrets. OpenAI says it has no interest in other companies’ trade secrets.

Apple has taken OpenAI to federal court. In a 41-page complaint filed July 10 in the U.S. District Court for the Northern District of California, first reported by Reuters, Apple accuses OpenAI, its chief hardware officer Tang Tan, former Apple engineer Chang Liu, and OpenAI’s hardware unit io Products of trade secret misappropriation and breach of contract. The case, Apple Inc. v. Liu, No. 5:26-cv-07078, is the most direct legal collision yet between the two companies, and it lands squarely on OpenAI’s consumer hardware ambitions.

What the complaint actually alleges

The complaint describes what Apple calls a “coordinated pattern of misconduct at an institutional level”: a scheme to extract confidential information about unreleased Apple products, engineering details, manufacturing and supply chain processes, vendor relationships, prototypes, CAD files, and internal codenames, all allegedly to accelerate OpenAI’s push into AI-powered consumer devices. Apple’s language is unusually aggressive for a company that normally litigates quietly, asserting that OpenAI’s hardware effort rests on a foundation that is “rotten to its core”.

The specifics are the striking part. Apple alleges that Chang Liu, a senior system electrical engineer who spent roughly eight years at Apple before joining OpenAI in January 2026, kept an Apple-issued laptop and exploited an authentication bug to download dozens of confidential files after his departure, including material on unreleased products. The filing says he also coached a prospective OpenAI recruit on evading Apple security and on which confidential materials to study before interviews.

The allegations against Tang Tan cut deeper because of who he is: a roughly 24-year Apple veteran who ran product design for iPhone and Apple Watch as a vice president, co-founded io Products with Jony Ive’s circle, and now runs hardware for OpenAI. Apple claims Tan used internal Apple codenames in interviews, directed job candidates to bring “actual parts” and prototypes from Apple for show and tell sessions, emailed himself supplier and industry information before leaving, and shared internal Apple security and departure procedures with new hires. Apple says it raised concerns with OpenAI in February 2026 and got no response.

Both sides, on the record

Apple’s prepared statement, carried by TechCrunch, says “significant evidence has emerged” that individuals employed by OpenAI wrongfully took confidential information about unreleased technologies, processes, and products, and that the company “will always defend our teams’ hard work and innovations”. OpenAI’s response, delivered through a spokesperson, was two sentences: the company has “no interest in other companies’ trade secrets” and remains focused on building its own technology. Apple is seeking an injunction barring use of its trade secrets, the return or destruction of materials, damages, and a jury trial.

The signal

Read past the legal claims and this is a fight about the next device category. OpenAI acquired io Products in 2025, in a deal widely reported at roughly $6.5 billion, precisely to build hardware that could displace the smartphone as the default AI interface. Apple, whose own AI story has been the weakest part of its narrative for two years, is now using its deepest institutional advantage (two decades of hardware process knowledge and the legal muscle to protect it) to slow that effort down. Discovery alone could force OpenAI to expose its device roadmap, suppliers, and timelines, which may be worth as much to Apple as any damages award.

It also fits a broader pattern we have tracked: as the AI race consolidates, the chokepoints are shifting from models to physical inputs and distribution, whether that is memory and power or, in this case, the tacit manufacturing knowledge needed to ship consumer hardware at scale. For OpenAI, which is simultaneously managing IPO speculation and an enormous capital program, a lawsuit that questions the provenance of its hardware program is a genuine reputational and diligence problem, not just a legal one.

What the framing hides

A complaint is an advocacy document, not a verdict. Every allegation above is exactly that: alleged, drawn from one side’s filing, untested by discovery or cross-examination. Employees moving between competitors is legal in California, which has long refused to enforce non-competes, and courts routinely distinguish between general skill and actual trade secrets. It is also fair to note what Apple’s framing leaves out: the company has lost a steady stream of senior hardware and AI talent to OpenAI and others, and litigation is a well-worn tool for making departures costlier. If Apple’s evidence of downloaded files and prototype requests holds up, this is serious misconduct. If it does not, the suit will read as an incumbent using the courts to slow a challenger it could not out-recruit.

What to watch

Three things matter from here. First, whether Apple moves for a preliminary injunction, which would force an early evidentiary showing and could disrupt io’s development schedule within months. Second, OpenAI’s formal answer: a motion to dismiss will signal confidence, while a quick settlement would suggest the document trail is real. Third, whether discovery drags Jony Ive’s LoveFrom or specific suppliers into the record, which would tell us how far along OpenAI’s device actually is. The first hardware war of the AI era has moved from the recruiting market to a San Jose courtroom, and the filings will tell us more about OpenAI’s device than any keynote has.

Get the Signal

AI and business tech news, verified by a physician who reads the filings. One email a week, no noise.

Dr. Joseph Joshua

Dr. Joseph Joshua is the founder and editor of Corewire. A medical doctor by training, he brings the evidence-first discipline of clinical medicine to technology journalism: claims get checked against primary sources before they get published. He has produced technology and B2B content for companies across…

View Bio

Keep Reading