Apple sued OpenAI, io Products, and two former Apple employees on July 10 in federal court in Northern California, accusing the AI company of using confidential Apple hardware information to accelerate its push into consumer devices.
The complaint, filed as Apple Inc. v. Liu, names Chang Liu, Tang Yew Tan, OpenAI Foundation, OpenAI Group PBC, and io Products as defendants. Apple says Liu, a former senior system electrical engineer, and Tan, a 24-year Apple veteran who is now OpenAI’s chief hardware officer, misused information tied to unreleased products, manufacturing processes, suppliers, and internal security procedures. The allegations have not been tested in court.
The case lands at a sensitive moment for both companies. OpenAI has been trying to move beyond software and cloud AI services into physical products, while Apple is still working to turn Apple Intelligence into a deeper product advantage across iPhone, Mac, Apple Watch, and its broader device lineup. A dispute that might otherwise look like a familiar employee-departure lawsuit is therefore also a fight over who gets to define the next consumer AI device.
What Apple alleges
Apple’s complaint says Liu left Apple in January 2026 to join OpenAI after eight years at the company. According to Apple, he failed to return an Apple-issued laptop, accessed Apple systems after leaving, and used what Apple describes as a previously unknown authentication bug to reach shared network folders. The filing says Apple fixed the bug after discovering it.
Apple also alleges Liu downloaded dozens of confidential hardware-related files while working for OpenAI, including engineering presentations, technical specifications, project data, and information about unannounced products. The complaint further claims Liu coached another Apple employee who was applying to OpenAI on what confidential Apple material to study before an interview and how to avoid trouble with Apple’s security team.
The claims against Tan are broader because of his role. Apple says Tan, formerly vice president of product design for iPhone and Apple Watch, used Apple project knowledge and supplier information while helping OpenAI build its hardware team. The complaint alleges he asked Apple job candidates about unannounced products and directed some candidates to bring Apple parts, design artifacts, or prototypes to interviews.
Apple also says OpenAI and io used confidential Apple information with suppliers, including in connection with a specific metal-finishing technique. The company is asking the court to block the defendants from using or disclosing Apple trade secrets, require the return of confidential materials, preserve evidence, and award damages.
OpenAI’s hardware plan is now part of the case
The lawsuit matters because OpenAI’s hardware work is no longer a side project. In 2025, OpenAI announced that the io Products team had merged with OpenAI, with Jony Ive and LoveFrom remaining independent while taking on design and creative responsibilities across OpenAI. The OpenAI announcement framed the combination as a push to create new AI-native devices rather than simply another app interface.
Apple’s complaint repeatedly ties the alleged misconduct to that hardware effort. It describes io as a venture co-founded by Tan and other former Apple leaders and says OpenAI acquired it to advance OpenAI’s entry into consumer hardware. That framing is central to Apple’s theory: this is not just about an engineer taking files, but about whether OpenAI’s device program benefited from Apple’s confidential playbook for designing, sourcing, manufacturing, and shipping hardware at scale.
OpenAI has denied interest in competitors’ secrets. In a statement reported by TechCrunch, the company said it remains focused on building technology that helps people and has no interest in other companies’ trade secrets. The litigation will now turn on evidence: device logs, server access records, interview practices, messages, supplier communications, and what OpenAI did after Apple says it raised concerns in February.
Why this is bigger than poaching
Employee mobility is normal in Silicon Valley, and Apple has often been on the other side of that dynamic. California law gives workers substantial freedom to change jobs, even to direct competitors. That makes Apple’s case depend less on the fact that OpenAI hired former Apple employees and more on specific alleged misconduct: retained devices, unauthorized access, copied documents, interview requests for protected material, and supplier use of confidential processes.
That distinction is important for AI companies racing into hardware. Frontier AI labs are hiring designers, silicon engineers, robotics specialists, device leaders, and supply-chain operators from established hardware companies. The competitive advantage is not only a model; it is also industrial design, thermal engineering, materials, component selection, factory process, test equipment, supplier negotiation, reliability data, and security controls around prototypes.
For OpenAI, the legal risk is unusually direct because the company is trying to build a new device category while recruiting from the company that built the most successful consumer device business of the past two decades. If Apple can show that confidential information moved into OpenAI’s hardware program, the remedy could reach beyond damages into restrictions on how OpenAI develops or ships a future device.
The Apple-OpenAI relationship is changing
The lawsuit also sharpens the breakdown in a relationship that recently looked cooperative. Apple brought ChatGPT into its software ecosystem as part of its 2024 Apple Intelligence rollout, while OpenAI gained a path into one of the world’s most valuable consumer platforms. But OpenAI’s hardware ambitions put it closer to Apple’s core business than a chatbot integration ever did.
If OpenAI’s first major consumer device is designed to reduce dependence on phone apps, screens, or the traditional mobile interface, Apple has a strategic reason to watch it closely. Apple’s own advantage rests on a tightly controlled stack: custom chips, industrial design, operating systems, App Store economics, accessories, services, and a manufacturing system that can ship devices globally at enormous scale. A serious AI-native device would need some version of that operational machinery, not just a better assistant.
The case is therefore a warning sign for the wider AI talent war. As AI companies move from models into devices, robots, cars, healthcare tools, and industrial systems, the line between hiring expertise and inheriting protected know-how will face more pressure. Companies recruiting from hardware incumbents will need cleaner interview rules, stricter onboarding checks, documented data-return processes, and stronger separation between what employees know generally and what they are contractually barred from using.
What to watch next
The next meaningful developments are likely to come from court filings, not social media reactions. Watch for whether OpenAI seeks early dismissal, whether Apple asks for emergency injunctive relief, and whether discovery reveals more about OpenAI’s device roadmap. Supplier evidence may matter as much as employee evidence, especially if Apple can show that a manufacturing process, material treatment, or partner relationship moved from Apple’s confidential environment into OpenAI’s hardware work.
The case could also force clearer public answers about OpenAI’s consumer hardware plans. So far, the company has said little about what it intends to ship, how close the product is, or whether it will compete with phones, smart speakers, wearables, or something less familiar. Apple’s lawsuit does not prove that OpenAI copied anything. It does, however, make OpenAI’s hardware effort harder to treat as vapor or speculation. Apple is now arguing in court that the effort is real enough, and threatening enough, to justify one of the most aggressive IP fights in the AI industry.
