Apple Sues OpenAI: The Battle Over Hardware Talent Poaching and Trade Secrets

From Partners to Courtroom Adversaries

Just two years ago, Apple and OpenAI were public partners—ChatGPT was integrated directly into Siri, becoming a key pillar of Apple's AI strategy. Today, however, Apple has formally filed suit against OpenAI, accusing it of systematic talent poaching and trade secret theft. The details of this lawsuit go far beyond a typical intellectual property dispute, striking directly at the most sensitive nerve in the war for hardware talent.

According to lawsuit details circulating on platforms like Reddit, Apple's core allegations center on two key individuals, against a broader backdrop: more than 400 former Apple employees now work at OpenAI. In its complaint, Apple bluntly states that the issues that have come to light so far are merely "the tip of the iceberg."

Core Allegation One: Tang Tan and the "Show and Tell" Interviews

The most prominent figure in the lawsuit is Tang Tan—OpenAI's current head of hardware. He was previously a vice president at Apple, having worked there for 24 years, and now leads OpenAI's hardware device strategy.

Apple alleges that Tang Tan made highly unusual arrangements during the recruitment process: he is accused of instructing current Apple employees interviewing at OpenAI to bring actual hardware components—including batteries, logic boards, and SIPs (System in Package)—to their interviews for physical demonstration.

SIP Technical Background: SIP (System in Package) is one of the core secrets of Apple's hardware strategy. Unlike traditional chip packaging, SIP integrates multiple independent chips—processor, memory, sensors, wireless modules, and more—into a single package, dramatically reducing size while improving energy efficiency. Apple's Apple Watch relies precisely on SIP technology to achieve extreme miniaturization, and its S-series chips are among the most complex SIP products in the industry. This kind of packaging design involves extremely high barriers in thermal management, signal integrity, and mechanical stacking processes. It is the crystallization of years of Apple's R&D accumulation, and one of the most difficult secret assets to reverse-engineer from public channels. This is precisely why the complaint singles out SIP by name—it represents not an ordinary component, but an indivisible competitive barrier.

Even more serious is the allegation that Tang Tan reportedly circulated to incoming OpenAI employees an internal Apple offboarding guide labeled "Need to Know," instructing them on how to circumvent Apple's security review process upon departure. If true, this crosses the line from normal talent movement into the organized subversion of a company's security mechanisms.

Core Allegation Two: Chang Liu and the Download of Confidential Cloud Files

The second key figure, Chang Liu, is a former Apple electrical engineer. The lawsuit claims that after joining OpenAI, he retained the laptop issued to him by Apple and discovered a system vulnerability—one that allowed him continued access to Apple's cloud network storage.

The complaint directly quotes his reaction at the time: "LOL, I found out I can still access [network storage], so funny." He then allegedly downloaded dozens of confidential files, many of which were explicitly marked "confidential."

This fully quoted conversation has become a devastating piece of key evidence in Apple's lawsuit. It not only proves the fact of unauthorized access, but also exposes the individual's lack of basic respect for trade secrets.

The "Smoking Gun" Mechanism in Trade Secret Litigation: Under the U.S. trade secret litigation system, the Defend Trade Secrets Act (DTSA) is the primary legal basis at the federal level, with states providing supplementary legislation (such as California's Uniform Trade Secrets Act). The outcome of trade secret cases often hinges on three core elements: the confidentiality of the information, the defendant's state of knowledge, and evidence of acquisition through "improper means." In the Chang Liu case, the directly quoted internal conversation precisely hits the latter two—it simultaneously proves that the individual knew he was accessing information without authorization (knowledge), and that this access was neither accidental nor in good faith (impropriety). Such instant message logs (IM logs) carry extremely high evidentiary weight in IP litigation, because they authentically capture the individual's subjective state of mind at the very moment of the incident—far harder for defense attorneys to challenge or reinterpret than after-the-fact statements.

Supply Chain Infiltration: Invoking Apple's Name

Beyond the allegations concerning talent and files, the lawsuit also touches on a sensitive aspect of the supply chain. Apple claims that OpenAI privately approached Apple's supply chain partners with the intent of using Apple's proprietary metal-finishing technique.

The crux of the problem: when approaching these suppliers, OpenAI allegedly falsely claimed that "Apple had granted a license"—when in fact Apple had never given any authorization.

The Moat of Hardware Manufacturing Processes: Metal-finishing techniques may seem like manufacturing details, but they are in fact among the hardest-to-replicate hidden barriers in tech hardware competition. Apple's signature aerospace-grade aluminum anodization, the sandblasted finish of its titanium bodies, and the mirror polishing of its stainless steel all require years of collaboration with specific suppliers to achieve mass-production yield and consistency. This kind of process know-how typically exists in the form of "tacit knowledge"—embedded in suppliers' operating procedures and technicians' experience—and cannot be fully conveyed through patent documents. For this reason, Apple and its supply chain partners often sign strict non-disclosure agreements prohibiting the provision of similar services to third parties. If OpenAI truly approached these suppliers under the guise of Apple's name, its essence was an attempt to use deceptive means to bypass the "moat" formed over years of supply chain development and achieve a technological leap.

If this allegation holds, its nature goes beyond mere internal secret leakage—it represents a direct erosion of Apple's core manufacturing process moat. The true competitive barriers in hardware often lie not in design blueprints, but in these hard-to-replicate mass-production processes and long-term supply chain relationships.

The AI Hardware Arms Race Reaches Fever Pitch

The timing of this lawsuit is intriguing. Just as Apple filed suit, it was reported that Apple is preparing to replace the original ChatGPT integration in Siri with Google's Gemini—former allies now fully at odds on both the AI and hardware fronts.

OpenAI's intent to build a hardware team has long been clear: acquiring Jony Ive's company io and aggressively recruiting Apple's core hardware talent both point to the same goal—building native AI hardware devices to challenge the personal computing paradigm defined by the iPhone.

Jony Ive and io: OpenAI's Hardware Ambitions: Jony Ive is the most influential Chief Design Officer in Apple's history, having led the design of a series of era-defining products including the iMac G3, iPod, iPhone, and Apple Watch. After leaving Apple in 2019, he founded the independent design firm LoveFrom. io is a hardware design company he co-founded with former Apple team members. In 2024, OpenAI announced its acquisition of io at a valuation of roughly $6.5 billion, a move widely interpreted as OpenAI's formal strategic declaration of entry into consumer hardware. The significance of this acquisition lies not only in gaining top-tier design capabilities, but even more in Jony Ive's deep network of connections accumulated across Apple's supply chain, manufacturing processes, and hardware ecosystem—which forms a highly sensitive intersection with the supply chain infiltration allegations in Apple's lawsuit.

Apple clearly views this development as a direct threat to the foundations of its hardware business. From a broader industry perspective, this lawsuit exposes the gray area of talent mobility in the AI era: when a company is eager to build hardware capabilities, the fastest path is to poach a competitor's most senior engineering teams en masse. But talent never takes only skills with it—it also carries inseparable confidential knowledge and proprietary experience. Apple's forceful move here is, in a sense, drawing a legal red line for talent competition across the entire tech industry.

One caveat: all of the above are allegations in Apple's one-sided complaint. OpenAI has yet to issue a formal response, and the ultimate facts remain to be determined by the court. But regardless of the outcome, this lawsuit signals that the AI hardware race has entered an entirely new phase—no longer merely a contest of products and capital, but a multi-front struggle involving talent, trade secrets, and the law.