OpenAI-Apple trade secrets lawsuit escalates

Apple's complaint reframes routine recruiting as a vector for trade-secret extraction. The filing claims OpenAI instructed candidates to bring prototypes and other 'design artifacts' for 'show and tell' sessions ^[1], turning an interview loop into a mechanism for pulling confidential hardware details out of Apple employees before they ever signed an offer letter. That pattern, Apple argues, was sustained rather than incidental: its original complaint alleges Chang Liu's work at OpenAI was informed by 'a steadily flowing stream of Apple's trade secret information' passed through Peng, and cites a message in which Liu bragged about retaining access to Apple's internal network storage, writing 'LOL, I found out I can access the [network storage], so funny' ^[2]. The detail that stands out is not a single leak but a system - interviews, retained credentials, and internal messaging all allegedly working together as a quiet data-collection loop around Apple's hardware division.

What began as a one-directional accusation has turned into dueling trade-secret claims. OpenAI now alleges that a former employee who returned to Apple transferred more than 100 confidential files, including manufacturing data and images of an unreleased product, to a personal account, and later briefed Apple's senior leadership on OpenAI's own product plans months before the current discovery fight became public ^[3]. OpenAI says it is 'very concerned that its trade secret and confidential information is now circulating inside Apple, where Apple engineers can shape Apple's upcoming products based on OpenAI's confidential and innovative work' ^[3]. Apple flatly denies this, saying it found no evidence to support the claim and has policies preventing new hires from using a prior employer's confidential information ^[4]. Meanwhile OpenAI, facing the prospect of an injunction, voluntarily agreed not to use Apple's alleged trade secrets while the case proceeds ^[5], and Apple's own attorney tried to narrow the scope of blame, arguing OpenAI doesn't have '400 bad apples,' but rather 'a few bad apples and know who they are' ^[5]- an attempt to keep the dispute contained to individuals rather than the company as a whole.

OpenAI's motion to dismiss does not just deny wrongdoing - it turns the lawsuit back on Apple's internal controls, arguing that Apple 'failed to identify any specific trade secrets that have been stolen' and that the suit 'actually exposes weaknesses in Apple's own security practices and human resources procedures' ^[1]. That framing lines up with why Judge Davila moved so quickly on discovery: the court cited a conversation between Liu and Peng about wiping Apple-owned electronics as grounds for expedited production, given the risk that forensic evidence could be destroyed ^[6]. It also matches how the story has landed publicly - on Reddit, commentary has focused on a Bloomberg-reported detail that Liu allegedly used a colleague's still-active Apple laptop credentials, combined with an unknown authentication bug, to access Apple's network after he had already left the company, fueling a running argument over whether Apple's own access-control failure mitigates the theft allegation, even as most commentary on Reddit remains sympathetic to Apple and skeptical of OpenAI's conduct.

Industry watchers are reading this less as an isolated spat and more as a preview of how AI companies will fight over physical-product talent. Tech analyst Paolo Pescatore argues the case shows competition shifting from software and models into factories and supply chains, predicting 'the next phase of the AI battle will be fought in factories and living rooms, not just in data centers' ^[7]. The practical risk for OpenAI is schedule, not just reputation: the litigation threatens to bog down its hardware plans even as it reportedly still hopes to announce a device this year and ship in 2027 ^[8].