Apple’s lawsuit accuses OpenAI of a pattern of secret-hunting around AI hardware push

Apple has sued OpenAI, alleging the ChatGPT maker recruited former Apple employees, accessed confidential materials, and tried to extract trade secrets tied to Apple’s hardware work as it prepares its own AI device. The 41-page complaint, filed this week, says the conduct threatens Apple’s product security and could shape the race to build the next generation of AI hardware.

The case centers on three former Apple staffers and a series of claims that, if proven, would paint a picture of unusually aggressive talent poaching and information gathering in one of tech’s most competitive new categories. Apple says the activity was not limited to hiring: it involved offboarding lapses, interview questions about unreleased projects, requests for physical components, and alleged coaching on how to avoid internal security procedures.

OpenAI has denied wrongdoing. A company spokesperson said it has no interest in other companies’ trade secrets and remains focused on building technology that helps users. Still, Apple’s filing is already drawing attention because it reaches beyond a normal employer dispute. It directly targets the methods a leading AI company may use while building hardware, and it raises fresh questions about how much the current AI boom depends on talent migrating across rival firms.

What Apple says is happening

Apple’s complaint describes what it calls a coordinated effort to gather proprietary information from inside the company and use it to support OpenAI’s hardware ambitions. The filing links that effort to OpenAI’s planned first consumer device, which is expected next year, and to the company’s expansion into product design and physical manufacturing.

According to Apple, the alleged conduct involves three former employees whose roles gave them access to sensitive engineering and product details. The company says those individuals did not simply change jobs; instead, they allegedly carried knowledge, documents, and interview tactics from one workplace to another in ways that violated Apple’s safeguards.

Who are the former Apple employees named in the case?

Apple’s filing names Tang Tan, Chang Liu, and Yu-Ting “Alyssa” Peng as central figures in the dispute. Tan spent 24 years at Apple and rose to vice president of Apple Watch before leaving in 2024 for Jony Ive’s hardware venture io, which OpenAI later acquired. OpenAI then made Tan its chief hardware officer.

Liu worked as a systems electrical engineer on the iPhone for more than eight years before joining OpenAI in January 2026 as a member of technical staff. Peng, another former Apple employee, joined OpenAI in April 2026.

Apple says the three were part of a broader pattern tied to OpenAI’s hardware development efforts. It alleges that the company’s hiring process and internal communications were used to identify, solicit, and leverage information about Apple products still under wraps.

How did Apple say confidential information was accessed?

Apple says one of the most serious allegations involves Liu’s departure from the company. In the filing, Apple claims Liu did not complete standard exit procedures, failed to respond to requests tied to his offboarding, and kept at least one Apple-owned computer after leaving. The company further alleges that he discovered a way to access Apple’s cloud storage weeks after departing and then used that access to download confidential files.

Apple says the material allegedly taken included technical specifications, unreleased product details, and engineering presentations about the company’s main logic boards. These are not trivial documents. Logic boards and related manufacturing processes sit at the center of product design, performance, and supply chain secrecy, which is why Apple treats them as highly sensitive.

In its complaint, Apple argues that Liu’s alleged access to network storage after departure was paired with messages suggesting excitement about being able to enter the system and retrieve materials that should have been off-limits.

Apple also says that Peng stayed in contact with Liu after his departure and shared Apple-related engineering information and vendor details while Liu was already working on OpenAI hardware. According to the filing, Liu allegedly advised Peng on how to copy files without drawing the attention of Apple’s security team and directed her to specific project folders and engineering data.

Why does Apple focus so much on offboarding?

Apple’s case is built in part on the idea that offboarding is one of the company’s last and most important security barriers. By its account, the moment an employee announces a departure is when a company can secure devices, revoke access, review downloads, and remind staff of confidentiality obligations.

Apple says some of the alleged misconduct happened precisely because those steps were not fully completed or were evaded. The company argues that this matters because trade secrets are easiest to lose when a departing worker still has partial access and a clear sense of where sensitive material is stored.

In practical terms, the lawsuit is not only about a few files. It is about whether a former employee can still navigate an employer’s systems after leaving, and whether another worker can continue feeding that person internal information from inside the company.

What OpenAI is accused of doing during hiring

Apple says OpenAI used interviews as an opportunity to probe for confidential details about unreleased Apple products. The filing claims Tan asked candidates to bring Apple components and product samples to interviews so they could be discussed in “show and tell” sessions, and that interview materials were used to surface information about Apple’s product engineering.

Apple also alleges that OpenAI asked applicants for more formal presentations that revealed highly sensitive information from their former work. In some instances, the company says, candidates reportedly prepared technical deep dives containing details that should not have left Apple’s walls.

The implication is that the interview process was not merely a hiring step, but a channel for extracting product knowledge from people with deep institutional memory. That raises a larger concern in Silicon Valley: when top engineers move between rivals, where is the line between a candidate describing professional experience and a candidate disclosing trade secrets?

What is ‘show and tell’ in this context?

In Apple’s telling, “show and tell” was not a harmless culture fit exercise. It was a request for former Apple staffers to bring physical parts and unfinished product elements to interviews so they could be examined by OpenAI personnel.

Apple says messages on one company-issued device show Tan directing an Apple employee to bring parts such as batteries, systems-in-package components, logic boards, and protective shields. According to the filing, he also suggested that these items could be useful for other interviewers to see.

Those are the kinds of objects that can reveal design priorities, materials, tolerances, thermal management, and manufacturing choices. In a hardware company, even a small component can say a lot about the broader product pipeline.

How does Apple say OpenAI prepared employees to avoid security checks?

Apple’s complaint goes beyond the alleged taking of information and says OpenAI also coached former Apple workers on how to avoid internal security scrutiny. According to the filing, Tan possessed a document outlining Apple’s offboarding procedures, and that information allegedly helped OpenAI advise departing Apple employees about what to expect.

Apple claims OpenAI told workers not to reveal their next employer, warned them about a possible “walk out” that would quickly remove them from the building, and advised them not to sign certain paperwork during exit interviews. In Apple’s view, those instructions were designed to help employees slip through a sensitive handoff without triggering extra review.

The company says it has seen a pattern in which staff leaving for OpenAI have supposedly taken steps to sidestep standard security measures. Apple says some employees ignored outreach from security personnel who were trying to schedule exit interviews and review procedures.

Apple contends in the filing that the alleged tactics appear to be working, pointing to what it describes as a recent rise in employees moving from Apple to OpenAI while avoiding normal departure checks.

What trade secret claims stand out most?

One of the most unusual parts of the case is Apple’s accusation that OpenAI tried to use Apple’s confidential information to access its suppliers. Apple says OpenAI approached a trusted partner that performs a specialized metal-finishing process used in Apple products and did so in a way that allegedly made the supplier believe it had Apple’s permission.

Apple argues it never authorized OpenAI or io to use its trade secrets or confidential methods, including proprietary knowledge passed to that supplier. The company also says OpenAI contacted at least one other manufacturer tied to Apple’s power and battery supply chain, using internal codenames and confidential details to ask targeted questions about components.

These allegations matter because hardware innovation rarely happens in isolation. It depends on specialized vendors, material scientists, and manufacturing partners who often work with the same companies for years. If a rival can move into that network armed with confidential information, the competitive damage could extend well beyond a single product design.

Why suppliers are so important in hardware disputes

Suppliers often know more about a product than outsiders realize. They may handle materials, finishes, tolerances, power systems, and assembly steps that never appear in public marketing materials.

That means trade secret disputes in hardware can involve far more than source code or internal memos. A confidential manufacturing process can be as valuable as a patent, especially when the goal is to launch a consumer device with a polished feel and precise industrial design.

Apple’s filing suggests it believes OpenAI was not only interested in what its former employees knew, but also in who made Apple’s products and how they were made. That would broaden the alleged misconduct from employee theft to ecosystem-level intelligence gathering.

A closer look at the timeline

The sequence of events Apple describes spans multiple years and several jobs changes. The company says the alleged conduct accelerated as OpenAI prepared to launch its own hardware line.

That timeline matters because it places the alleged conduct in the context of a broader hardware race. OpenAI is no longer just a software company delivering chatbot responses. It is now trying to build physical products, and Apple is one of the few companies in the world that has spent decades perfecting the hardware business at scale.

How serious is this legally?

This lawsuit could become a significant test of how courts treat trade secret disputes in the age of AI hardware. If Apple can show that former employees retained access, copied confidential materials, and used internal knowledge to assist a rival, it may strengthen the legal case around misuse of proprietary information.

But the burden of proof is high. Trade secret claims typically require evidence that the material was genuinely confidential, that reasonable steps were taken to protect it, and that the defendant improperly acquired or used it. OpenAI is expected to argue that it hired people for their expertise, not for Apple’s private data.

There is also a reputational dimension. Even if the case resolves without a finding of liability, the allegations may influence how companies screen new hires, handle offboarding, and police what candidates can discuss in technical interviews.

What could the case mean for AI hiring practices?

If the lawsuit gains traction, it could push AI firms and hardware startups to tighten candidate interviews and internal compliance systems. Employers may become more cautious about asking candidates to bring previous work samples, describe unreleased projects, or share presentation decks tied to prior jobs.

It could also make departing workers more careful about what they keep on old devices, what access they retain after resignation, and how much they discuss a new employer before their exit is complete.

For the broader industry, the case is a reminder that the AI race is no longer only about models, chips, and data centers. It is also about product design, supply chains, and the fragile boundary between talent transfer and corporate espionage.

What OpenAI says in response

OpenAI has rejected the accusation that it is interested in stealing competitors’ secrets. In a statement provided to The Verge, the company said it does not want other companies’ trade secrets and remains focused on making technology that helps people everywhere.

That response is likely to be the starting point for a broader legal fight. The company will need to address not only the conduct of named employees, but also Apple’s claims about interview practices, supplier contact, and internal coaching around security procedures.

For now, the complaint is only one side’s version of events. Still, the filing is detailed enough to suggest Apple believes it has a substantial paper trail, including devices, messages, and system records. That makes the dispute unusually combustible for two of the most closely watched names in tech.

Why this case matters beyond Apple and OpenAI

This is bigger than a company fight over a few engineers. The lawsuit could help define the rules for how AI companies build physical products while competing for veteran hardware talent from incumbents like Apple.

There are three reasons the case has broader significance:

• It tests the limits of employee mobility. Tech workers frequently move between rivals, but this case asks how much prior knowledge is too much.
• It spotlights hardware, not just software. AI competition is increasingly about devices, suppliers, and industrial design.
• It raises compliance expectations. Companies may be forced to document offboarding and interview procedures more tightly.

Apple, of course, has a long history of protecting its product secrecy with unusual intensity. OpenAI, meanwhile, has been moving rapidly from a software-first company into one that may soon sell its own branded hardware. Those paths were likely to cross sooner or later.

Key allegations at a glance

Here is a simplified view of the claims Apple is making in the case.

What happens next?

The next stage will likely focus on whether Apple can substantiate the technical and documentary claims outlined in its complaint. Discovery could become contentious if the case proceeds, because both companies may seek internal communications, device logs, interview records, and supplier correspondence.

The legal fight may also turn on how much of the alleged information can be shown to have been confidential, whether OpenAI used it directly, and whether any staffers violated written obligations when they changed employers. Those details will determine whether this becomes a high-profile but narrow employment dispute or a landmark case about AI-era trade secrets.

For now, the filing adds one more layer of tension to an already intense rivalry. Apple is trying to protect the secrecy that has long defined its hardware empire. OpenAI is trying to build the kind of physical product that could reshape its business. The lawsuit suggests those ambitions may now be colliding in court.

As the industry watches, one thing is clear: the fight over AI hardware is no longer theoretical. It has moved from product announcements and talent headlines into accusations of stolen know-how, supplier deception, and security evasion — the kind of claims that can define how the next generation of tech companies compete.