The Allegations: Apple Points Finger at OpenAI
A simmering conflict between Apple and OpenAI over the future of AI hardware has boiled over into public view, according to a report from The Verge. The dispute centers on the departure of two key engineers from Apple to OpenAI, including Tang Tan, who now serves as OpenAI's chief hardware officer, and Chang Liu, who joined the AI company from Apple in January 2026. The report also references a hardware startup Apple acquired in 2025, suggesting that the talent exodus is tied to a specific project or technology that Apple views as proprietary.
While the legal standing of any potential action remains unclear, the very public airing of Apple's concerns indicates that the company sees OpenAI's hiring spree as more than just aggressive recruiting—it views it as a direct threat to its secretive hardware roadmap. The report, originally surfaced by 9to5Mac, has sparked debate in the tech community about the limits of talent mobility in an era where AI and consumer hardware are converging faster than ever.
Tang Tan and Chang Liu: Key Architects of Apple's Hardware Vision
Tang Tan's move to OpenAI is especially notable. As a longtime Apple veteran, Tan was heavily involved in the development of several iconic products. His appointment as OpenAI's chief hardware officer signals a serious commitment by the AI giant to build physical devices, not just software models. Chang Liu, who joined OpenAI in January, is another critical hire; her background reportedly includes work on advanced sensor systems and miniaturized components—assets that would be invaluable to any company aiming to create an AI-first consumer device.
The 2025 Apple acquisition mentioned in the report likely refers to a startup specializing in either advanced displays or low-power edge AI processing chips. If Tan and Liu were instrumental in integrating that acquired technology into Apple's pipeline, their departure to a direct competitor like OpenAI raises legitimate questions about the protection of trade secrets. Under California law, non-compete agreements are largely unenforceable, but trade secret laws remain a powerful tool for companies like Apple to litigate against former employees who might inadvertently or deliberately carry proprietary knowledge to a rival.
Apple's Statement: "Breakthrough Technologies" and a Veiled Warning
Apple, typically tight-lipped about internal matters, broke its silence with a statement shared via 9to5Mac: "At Apple, our teams are constantly developing breakthrough technologies to create the best products in the world. We take the protection of our intellectual property seriously, and we will defend it vigorously when necessary." Although the statement does not name OpenAI outright, the context makes the target clear. The language—"defend it vigorously"—is a chosen escalation from Apple's standard boilerplate, suggesting that behind the scenes, legal teams are preparing or already engaged in a dispute.
What remains unknown is whether Apple has filed a formal complaint or is merely issuing a warning shot. Legal experts note that for a trade secret claim to stick, Apple would need to demonstrate that the employees took tangible information—code, schematics, or documented processes—rather than simply relying on general knowledge. The burden of proof is high, but the chilling effect on OpenAI's ability to rapidly build out a hardware division could be an outcome Apple welcomes regardless of litigation success.
The Legal Landscape: Trade Secrets and Non-Competes in California
The backdrop to this dispute is California's famously employee-friendly legal environment. Non-compete clauses are void by default, reflecting a public policy that encourages mobility and innovation. However, trade secret misappropriation under the Defend Trade Secrets Act (DTSA) or California's Uniform Trade Secrets Act remains a potent threat. The departure of multiple employees with overlapping expertise in a specific domain—coupled with the timing of an acquisition—can create a pattern that courts view as suspicious.
Historical parallels exist: the Waymo vs. Uber case involved allegations that a former engineer downloaded thousands of files before leaving. There is no evidence yet that such behavior occurred here, but the mention of a 2025 acquisition in the reporting implies Apple may be trying to establish that a critical mass of confidential knowledge migrated to OpenAI in a short period. For OpenAI, the legal risk is twofold: potential liability and the possible injunction that could slow down its hardware efforts just as the company is rumored to be developing an AI wearable or assistant device.
What This Means for the AI Hardware Race
This spat is a harbinger of the intensifying talent wars in the AI sector. As companies like OpenAI expand from pure software into hardware, they inevitably collide with established players like Apple, which has one of the most sophisticated silicon and design teams in the world. The fight over Tang Tan and Chang Liu reflects a broader struggle: who will define the next generation of AI-native devices? Apple has the supply chain, the brand, and the design legacy. OpenAI has the models and the user base from ChatGPT, but it lacks hardware expertise, which it is racing to build by hiring away the very people who built Apple's success.
For developers and tech professionals, the case underscores the precarious balance between career growth and legal exposure. Working on cutting-edge AI hardware often means carrying knowledge from one company to the next, but as this dispute shows, the definition of "general skill" versus "trade secret" is increasingly contested. The coming weeks will reveal whether Apple escalates to official legal action or if the public naming of individuals serves as a deterrent. Either way, the message is clear: the AI hardware race is not just about algorithms and chips—it's a battle over the minds building them, and Apple is drawing a line in the sand.
评论