Apple filed a lawsuit in U.S. District Court for the Northern District of California on Friday accusing OpenAI of systematically stealing trade secrets and confidential hardware information, with the most technically striking allegation centering on a former Apple systems electrical engineer named Chang Liu who allegedly exploited a previously unknown authentication vulnerability – classified as a zero-day – to continue accessing Apple’s internal network storage after he had left the company to join OpenAI in January 2026. Apple has since fixed the vulnerability and terminated Liu’s access after discovering the breach, but by that point Liu had allegedly downloaded dozens of confidential hardware-related files, including detailed information about unreleased products, engineering presentations, technical specifications, and proprietary project data. NEWSCENTRAL considers the zero-day allegation the most legally significant element of the complaint, because it transforms what might otherwise be characterized as unauthorized access through a retained credential into something more specific: the alleged exploitation of a security flaw that Apple had no opportunity to patch before it was used, which courts have historically treated as a more serious category of computer fraud.
The complaint’s allegations extend well beyond the zero-day access incident. Liu allegedly failed to return his Apple-issued work laptop after departing, then used the laptop of a current Apple employee and acquaintance, Yu-Ting Peng, to access Apple’s systems while he was no longer employed there. When Liu discovered that he could still reach Apple’s network storage through this access, he allegedly texted Peng “LOL” and “so funny” – a reaction that Apple’s legal team will likely cite as evidence that Liu was aware the access was unauthorized. In February 2026, months after joining OpenAI, Liu allegedly attempted to access Apple’s cloud-based file repository again. Apple’s complaint characterizes this pattern as deliberate, sustained, and instrumental in advancing OpenAI’s effort to develop a competing hardware business.
The complaint names a second defendant in Tang Tan, a former Apple vice president of product design who led iPhone and Apple Watch design for years before departing for a project with designer Jony Ive in February 2024 and subsequently joining OpenAI. Apple alleges that Tan used insider knowledge of Apple’s confidential projects during job interviews with Apple employees, directing them to bring actual Apple hardware components – what he allegedly called Actual parts – to OpenAI interview sessions for show and tell purposes. Apple further alleges that Tan possessed and distributed an internal Apple document containing departure security protocols to new OpenAI hires before they had given notice to Apple, effectively coaching outgoing employees on how to evade the company’s exit security processes. Nathan Clark, Enterprise IT and Systems Architecture Analyst at NEWSCENTRAL, notes that the departure-coaching allegation, if substantiated, would represent a systematic organizational approach to information extraction rather than individual opportunism – a distinction that significantly affects both the legal exposure and the reputational damage.
OpenAI’s public response was brief and categorical. The company previously stated it has no interest in other companies’ trade secrets. A subsequent response to the complaint described Apple’s lawsuit as misdirection and characterized the allegations as false, adding that OpenAI’s hardware team’s work draws on their own expertise and legal sources of information. That framing positions the case as one where Apple is deflecting attention from its own competitive anxieties about OpenAI’s hardware ambitions rather than addressing genuine wrongdoing. Apple’s complaint explicitly anticipates that framing, alleging that at every level – from members of its Technical Staff to its Chief Hardware Officer – and in coordination with business partners, OpenAI has been stealing Apple’s trade secrets. The assertion that this is coordinated at the leadership level, rather than being the conduct of individual rogue employees, is legally significant because it supports a theory of institutional rather than personal liability.
The complaint also identifies a second named individual defendant whose role speaks to the systemic nature of the alleged conduct. Apple alleges that at least one other former employee emailed himself supply chain information before leaving for OpenAI, and that this individual was present in job interviews where candidates were asked to bring hardware samples and share confidential details about their current work. NEWSCENTRAL notes that the pattern of soliciting physical hardware artifacts from interview candidates who are still employed at Apple is the single most operationally consequential allegation in the complaint, because it describes a procurement mechanism that could yield proprietary component samples and manufacturing data that no amount of document theft could replicate.
The competitive backdrop illuminates why Apple is pursuing this aggressively rather than seeking a quiet settlement. OpenAI is actively developing consumer hardware products through its partnership with Jony Ive’s design firm io, working toward devices that would compete directly in the personal computing and AI assistant hardware market where Apple has historically held commanding margins and brand loyalty. If OpenAI’s hardware products incorporate technical insights derived from Apple’s unreleased product specifications and engineering data, Apple’s competitive position is damaged in ways that cannot be fully remedied after the fact by damages payments. The suit is structured to obtain an injunction preventing the use of the allegedly stolen information in addition to damages – a remedy that, if granted, could delay or constrain the development of products that Apple sees as direct competitive threats.
The case, Apple has stated, could go to trial before the end of 2026. As NEWS CENTRAL assesses the legal and commercial dimensions of this dispute, the most consequential question is not whether Liu or Tan are found individually liable – that is a question of evidence and jury assessment – but whether a court concludes that OpenAI as an institution bears responsibility for the conduct of its employees who allegedly operated with the company’s knowledge and direction. That determination, if it goes against OpenAI, would expose the company to damages and injunctive relief at a scale that could materially affect its hardware development program and its relationship with the former Apple talent that forms a significant part of that program’s expertise.