Apple accuses OpenAI of destroying evidence in trade secrets case over stolen circuit design
Apple has formally accused OpenAI of destroying evidence in a trade secrets lawsuit, alleging former engineer Chang Liu downloaded a confidential power-converter schematic and used it within OpenAI's AI workflow. Apple claims Liu directed a colleague to destroy evidence after learning of the investigation. OpenAI calls the dispute "a mess of Apple's own making." Apple seeks monetary damages, an injunction, and destruction of proprietary materials. A judge will hear arguments on October 1.
Editorial responsibility
- No named human review is recorded for this page.
- Reports are grouped by semantic similarity and deterministic rules. Language models may assist titles, summaries, translation and cross-source analysis; the page itself is projected from evidence records.
- Current automated evidence projection
Cross-source coverage
Common ground
- The evidence destruction by the employee is the strongest part of Apple's case and suggests guilt.
- The case will likely survive summary judgment due to the evidence destruction.
- Model interrogation could provide a smoking gun if it shows verbatim reproduction of Apple's circuit values.
- The case is about proof of specific trade secret use, not a sweeping precedent for all AI training.
Points of contention
- Whether AI training creates a fundamentally new legal problem or just a complex damages calculation.
- Whether 'statistically indistinguishable performance' is enough for a jury to connect stolen data to model behavior.
- Whether the absence of memorization in the model proves innocence or just shows data was digested.
- Whether this case will set a broad precedent for AI accountability or remain a narrow dispute about employee misconduct.
Blind spots
- Neither side fully addressed how companies can watermark proprietary data to survive model training.
- The debate overlooked the possibility that Apple's schematics were already public through patents or papers.
- The technical question of model memorization versus generalization was raised but not deeply explored with concrete examples.
WorldAttention’s read
This case hinges on proof, not principle. Apple has a strong narrative with evidence destruction and employee behavior, but a weaker technical case unless they can show the model reproduced specific secret values. The judge will likely let the case proceed, but Apple will struggle at trial without a clear fingerprint from model interrogation. The real takeaway is that companies need to start watermarking data to track it through AI training, rather than relying on outdated legal frameworks.
Wire timeline
Apple accuses OpenAI of destroying evidence in trade secrets case
Apple has filed a court brief accusing OpenAI of destroying evidence in their ongoing trade secrets dispute. The filing, submitted in support of Apple's motion for expedited pretrial fact-finding, centers on a forensic analysis of an Apple-issued MacBook used by former Apple engineer Chang Liu after he joined OpenAI. Apple claims the analysis found that Liu and others at OpenAI were aware of his continued unauthorized access to Apple's third-party cloud storage providers. Apple also alleges that after Liu learned he was under investigation, he directed OpenAI colleague Yu-Ting Peng in June to destroy evidence, and that Peng agreed to do so. OpenAI filed its own brief calling the dispute 'a mess of Apple's own making' and arguing that any Apple documents Liu accessed were irrelevant to his work at OpenAI. The two companies have been in court since Apple sued OpenAI in July, accusing it and two former employees of misappropriating trade secrets to advance OpenAI's push into consumer hardware. Apple is seeking monetary damages, a preliminary injunction blocking OpenAI from working on hardware based on Apple's technology, and an order requiring the destruction of any proprietary materials. A judge is scheduled to hear arguments on the injunction motion on October 1.
Apple says former engineer used its confidential circuit design inside OpenAI's AI workflow
In a new legal filing reported by Reuters, Apple alleges that former electrical engineer Chang Liu downloaded a confidential power-converter schematic after leaving the company and used it within OpenAI's AI workflow. According to the filing, Liu ran the schematic in LTspice on a Mac mini, generating simulation outputs. Apple claims his AI agent learned to operate the LTspice simulator, inspect results, and tune compensation parameters, reducing his stated workflow from a day to two hours. Apple argues this raises a distinct AI risk because proprietary information may become harder to trace or remove after an AI agent learns from it. The filing does not establish that OpenAI trained model weights on Apple's data, but describes an AI agent learning to operate the simulator and tune parameters. The court has not ruled on trade secret theft; Apple submitted this evidence to support its request for expedited discovery.
Apple accuses OpenAI of destroying evidence in trade secrets case
In a new court filing, Apple has accused OpenAI of actively destroying crucial evidence related to a trade secrets case. The filing alleges that former iPhone engineer Chang Liu downloaded a confidential Apple circuit schematic and subsequently used it in his work at the AI company. This legal development highlights ongoing tensions between major tech firms over intellectual property and employee mobility, particularly as AI companies aggressively recruit talent from competitors. The case underscores the high stakes involved in protecting proprietary technology in the rapidly evolving AI sector.
Show 2 older updatesHide older updates
Apple Accuses OpenAI of Destroying Evidence in Trade Secrets Lawsuit
Apple Inc. has accused OpenAI of destroying evidence in a trade secrets lawsuit, according to a report from Bloomberg. The allegation, shared via a post on X by the financial news outlet Unusual Whales, marks a significant escalation in the legal dispute between the two technology companies. Apple's claim suggests that OpenAI may have intentionally disposed of materials relevant to the ongoing litigation. The specific details of the trade suit and the nature of the evidence allegedly destroyed have not been fully disclosed in the initial report. This development adds a new layer of contention to the legal battle, which centers on proprietary information and intellectual property. The accusation, if proven, could have serious implications for OpenAI's legal standing and the broader case. Neither Apple nor OpenAI has issued an official statement regarding the claim at this time. The case is being closely watched by the tech industry as it involves two major players in the artificial intelligence and consumer electronics sectors.
Apple accuses OpenAI employee of accessing secret circuit plans and training AI agent
In a breaking development, Apple has formally accused an employee of OpenAI of accessing confidential power converter circuit plans and using that proprietary Apple data to train an AI agent. The accusation, reported via Polymarket, alleges that the OpenAI employee improperly obtained secret design documents related to Apple's power converter technology. The employee is then said to have leveraged this stolen intellectual property to train an artificial intelligence model, raising serious concerns about data security and corporate espionage between two major tech companies. The incident highlights the growing tensions over proprietary data in the AI industry, where access to unique datasets can provide a competitive advantage. Apple has not yet released an official statement beyond the accusation, and OpenAI has not publicly responded to the claims. The case could have significant legal and reputational implications for both firms, as well as broader implications for how AI companies handle sensitive third-party data.