Apple and OpenAI have spent the past few years presenting themselves as partners in the AI era. Now, that relationship has fractured in spectacular fashion. Apple has sued OpenAI, its hardware unit io Products, and two former Apple executives, alleging a deliberate campaign to take confidential information tied to unreleased devices, manufacturing processes, supplier relationships, and other closely guarded parts of Apple’s hardware operation.

The case, filed in federal court in San Jose, California, goes far beyond the usual Silicon Valley argument over a departing employee taking a few files on the way out. Apple describes what it calls a “pattern of theft” involving former staff who allegedly accessed, copied, solicited, or used confidential company material as OpenAI raced to build a new category of AI hardware. The company has named OpenAI chief hardware officer Tang Tan and former Apple employee Chang Liu, alongside OpenAI and io Products, in the complaint.

At the center of the dispute is an uncomfortable question for the AI industry: when a company’s next big product category is still mostly invisible to the public, how much of its advantage lies in the people it hires – and where is the line between recruiting talent and acquiring the knowledge they were contractually bound to leave behind?

Apple’s lawsuit argues that OpenAI crossed it.

According to the complaint, Liu allegedly accessed Apple systems after his employment ended and downloaded dozens of confidential files. Apple says those materials included detailed information on unreleased hardware, internal engineering presentations, technical specifications, and proprietary project data. The company also alleges that Liu advised a former colleague on how to copy Apple files and evade the company’s security teams before joining OpenAI, including using the Line messaging app to communicate. These are allegations, not facts established by a court, but their specificity signals that Apple intends to litigate aggressively.

The allegations involving Tan are broader. Apple claims that its former executive emailed himself confidential supplier information before leaving and later sought confidential details while interviewing Apple staff for OpenAI. The complaint says OpenAI personnel encouraged prospective hires to bring materials such as prototypes and CAD or design artifacts to interviews. Apple’s central argument is that OpenAI was not merely competing for employees – it was allegedly trying to obtain the hard-won industrial knowledge behind Apple’s products.

That distinction matters. Tech companies hire from one another constantly, especially in a region as tightly interconnected as Silicon Valley. Engineers change jobs. Designers bring experience. Executives carry instincts, management habits, and a mental map of how an organization works. None of that is inherently improper. Trade-secret law is meant to protect the more concrete and confidential layer underneath: non-public technical documents, supplier arrangements, manufacturing methods, pricing, product roadmaps, prototype details, and other information that creates a competitive edge.

Apple has built its business around precisely that kind of control. Its advantage has never been limited to industrial design or software. It includes a tightly managed global supply chain, unusually close relationships with component makers and manufacturers, and a years-long ability to turn difficult engineering decisions into polished consumer products at enormous scale. The company’s lawsuit specifically alleges that OpenAI sought access to Apple’s partner network, including an industrial-design and metal-finishing partner that Apple says was asked to perform proprietary processes for OpenAI’s benefit.

That claim gets to the real stakes of the case. OpenAI does not need Apple’s secrets to make a chatbot. But it does need a different kind of expertise if it wants to create a mainstream physical device that people will carry, wear, or use every day. That includes hardware design, manufacturing yields, materials, component sourcing, prototyping, reliability testing, packaging, and the messy logistics of actually shipping millions of devices.

OpenAI’s interest in that world became much clearer when it acquired io Products, the hardware startup associated with former Apple design chief Jony Ive, in a deal valued at roughly $6.5 billion. The transaction brought OpenAI much closer to the consumer-device business and gave it access to a team with deep Apple roots. OpenAI has said io formally merged with the company in July 2025, while Ive and his LoveFrom design firm remained independent.

Tang Tan’s presence made that hardware push even more significant. At Apple, Tan was closely associated with product design and hardware execution. For OpenAI, he represents something the company has historically lacked: seasoned operational leadership for turning ambitious AI ideas into physical products. Apple’s complaint presents that experience as inseparable from the confidential information it says was improperly taken and used.

OpenAI, for its part, has rejected the premise of Apple’s claims. Drew Pusateri, an OpenAI spokesperson, told The Verge that the company has “no interest in other companies’ trade secrets” and remains focused on building technology that empowers people. That response is concise, but it places the dispute on a familiar legal track: Apple will need to show that identifiable trade secrets existed, that reasonable steps were taken to protect them, and that the defendants improperly acquired, disclosed, or used them.

The timing makes the lawsuit especially awkward. Apple and OpenAI have not been simple rivals. Apple integrated ChatGPT into parts of its broader AI strategy, positioning OpenAI as a partner that could help users access outside expertise through Apple’s software ecosystem. But the lawsuit illustrates how quickly that alignment can weaken when both companies see the same future opportunity.

For Apple, that opportunity is AI hardware. For OpenAI, it may be even more urgent. The company became one of the most influential names in technology through software, but its leaders have repeatedly suggested that generative AI could require new kinds of computing devices rather than simply living inside smartphones and laptops. The industry has already watched multiple companies attempt that leap, with mixed results. An AI-first device has to be compelling enough to create a new habit, yet useful enough that consumers do not simply reach for the phone already in their pocket.

OpenAI is expected to introduce its first hardware product next year, according to the reporting cited in Apple’s complaint. Apple, unsurprisingly, frames that impending launch as the reason the alleged conduct matters. Its suit argues that OpenAI’s hardware business has leaned on work that took Apple decades to develop.

There is also a human-resources angle that could make the case resonate well beyond these two companies. Apple claims that more than 400 of its former employees now work at OpenAI. On its own, that number does not prove anything improper. OpenAI is one of the best-funded and most aggressively expanding technology companies in the world, and Apple remains one of Silicon Valley’s largest employers. Employee movement between the two is inevitable.

Still, Apple alleges that OpenAI advised departing employees to report if Apple asked them to sign anything. If substantiated, that detail could become particularly important because it suggests the dispute was not merely about individual judgment calls by former workers, but about company-level conduct. OpenAI will likely argue that recruitment and competitive hiring are legitimate, and that any alleged misconduct by individuals should not automatically be attributed to the company.

For now, this is a complaint full of claims that remain unproven. Apple will seek to turn those claims into evidence. OpenAI and the individual defendants will have the opportunity to challenge both Apple’s account and its definition of what constitutes a trade secret. Discovery could be the most consequential phase, exposing internal communications, recruiting practices, hiring materials, device-development documents, and the extent to which Apple’s allegedly confidential information appeared inside OpenAI’s operations.

The broader story, though, is already clear. AI’s next battle may not be fought only in model benchmarks, data centers, and subscription tiers. It may also be fought in design studios, prototype labs, factory floors, and supply-chain negotiations – the less glamorous parts of technology that determine whether an idea becomes a real product.

Apple has spent decades mastering that machinery. OpenAI is trying to build it at remarkable speed. This lawsuit suggests Apple believes the race has become personal.