OpenAI has responded publicly to Apple’s trade secrets lawsuit by releasing private emails and messages that it says challenge Apple’s account of how the dispute unfolded.
Apple asked a court on August 3 for a preliminary injunction that would prevent OpenAI from using or disclosing information the company describes as proprietary and confidential. OpenAI answered through a blog post titled “Apple is getting this wrong,” publishing correspondence from Apple’s outside legal counsel and iMessages involving Chang Liu, a former Apple engineer who now works for OpenAI.
The released material does not directly resolve the central legal question: whether OpenAI or its employees improperly obtained or used Apple’s trade secrets. Instead, it supports OpenAI’s broader argument that Apple handled the matter carelessly, failed to communicate its concerns clearly and is presenting Liu’s conduct in an unfairly suspicious way.
The strategy moves part of the dispute beyond the courtroom and into public debate. That may influence how the companies are perceived, but it also creates risks for OpenAI if the disclosures appear selective or unrelated to the core allegations.
Apple Seeks a Preliminary Injunction
Apple filed its lawsuit on July 10 and later asked the court for a preliminary injunction against OpenAI.
The requested order would bar OpenAI from using or disclosing information Apple identifies as its proprietary and trade secret material.
A preliminary injunction can impose restrictions before a case reaches a final judgment. Apple therefore appears to be seeking immediate protection while the wider dispute continues.
The source does not specify every category of information Apple believes was taken or how OpenAI allegedly used it. The core accusation is that confidential hardware-related material was improperly transferred through Liu after he left Apple.
OpenAI denies that it possesses or wants Apple’s trade secrets.
OpenAI Releases Emails and iMessages
Instead of limiting its response to legal filings, OpenAI published private communications.
The company released emails involving Apple’s outside counsel and iMessages connected to Liu.
OpenAI appears to be using the messages to argue that Apple’s lawsuit is “careless, aggressive and oddly personal.” It also says it does not possess and does not want Apple’s trade secrets.
The disclosures are intended to weaken Apple’s broader narrative by questioning the company’s legal process and its description of Liu’s conduct.
However, the messages do not by themselves establish whether confidential information was taken, whether that information qualified as a trade secret or whether OpenAI used it.
They address credibility and context more directly than the underlying legal claim.
The Email Mix-Up Raises Questions About Apple’s Outreach
One of OpenAI’s main points concerns an apparent communication error by Apple’s lawyer.
According to the released correspondence, Apple’s outside counsel contacted the wrong person at OpenAI and later thanked that individual for a telephone conversation that had never occurred.
The incident had already become public before OpenAI released the latest materials.
OpenAI appears to be arguing that Apple cannot fairly claim that the AI company ignored its concerns when the initial outreach was mishandled and no effective follow-up discussion took place.
This may support OpenAI’s claim that Apple escalated the dispute without properly presenting the specific allegations first.
Still, a legal communication error does not determine whether the underlying trade secret claims are valid.
Apple’s lawyer contacting the wrong person may reflect poor case management, but it does not prove that no confidential files were transferred or used.
The Messages Challenge the Description of Liu’s Conduct
Apple accuses Liu of secretly downloading dozens of confidential files after leaving the company.
The iMessages released by OpenAI appear to show Apple employees asking Liu for help accessing files, including material linked to his former work iCloud account.
OpenAI seems to be using this exchange to challenge the characterization that Liu acted entirely in secret.
If Apple employees knew he had access to certain files and contacted him for assistance, that context could complicate the claim that every interaction with the material was concealed.
However, the messages do not necessarily answer several important questions.
They do not establish whether all downloaded files were included in those requests, whether Liu had permission to retain them or whether the material later reached OpenAI.
The communications may weaken one aspect of Apple’s presentation without resolving the broader allegation.
The Core Trade Secrets Claim Remains Unanswered
The central issue is not whether Apple’s lawyers made mistakes or whether some employees contacted Liu after his departure.
The central issue is whether protected company information was improperly taken and whether OpenAI coordinated, encouraged or benefited from that conduct.
The published messages do not directly disprove that allegation.
They may help OpenAI portray Apple as careless or overly aggressive, but they do not provide a complete factual record of what files existed, where they were stored, who accessed them or whether they influenced OpenAI’s work.
Those questions are more likely to be addressed through formal discovery, technical analysis, testimony and court filings.
The public disclosures therefore function mainly as a reputational response rather than a comprehensive legal defense.
OpenAI Is Fighting in the Court of Public Opinion
By publishing private communications, OpenAI is attempting to shape the public interpretation of the lawsuit.
This strategy can be valuable in a high-profile corporate dispute. Investors, employees, potential recruits and customers may form views before a court reaches any conclusion.
OpenAI may also believe that portraying Apple’s process as disorganized will reduce the credibility of the wider case.
Public opinion could matter even more if Apple receives the jury trial it requested. Jurors are instructed to decide cases based on admitted evidence, but widespread public narratives can still affect the environment surrounding litigation.
The strategy is not without danger.
If the released messages are seen as selective, incomplete or designed to distract from the central allegation, they may strengthen criticism of OpenAI rather than weaken Apple’s position.
Publishing Private Messages Creates Its Own Risks
OpenAI’s decision introduces questions about whether private correspondence should be used as a public-relations tool during active litigation.
The company may argue that publication was necessary to correct a misleading account.
Critics may view it as an attempt to embarrass Apple’s lawyers and employees without addressing the substantive accusation.
Selective publication also creates an information imbalance. The public sees the messages OpenAI chose to release, but not necessarily the full context surrounding them.
Additional communications could support a different interpretation.
That uncertainty limits how confidently the materials can be assessed outside the legal process.
OpenAI Has Used This Tactic Before
This is not the first time OpenAI has released private correspondence during a legal dispute.
In 2024, the company published emails from Elon Musk while responding to his lawsuit alleging that OpenAI had abandoned its original nonprofit mission.
Musk later withdrew that case before filing a similar action several months afterward.
The earlier episode shows that OpenAI views direct publication of internal or private communications as a legitimate defensive strategy.
Such releases can generate headlines and shift attention toward inconsistencies in an opponent’s claims.
They do not necessarily end litigation, however, and may simply create another layer of conflict around the underlying dispute.
A Court Order Could Expose More OpenAI Records
The immediate legal risk for OpenAI is Apple’s request for a preliminary injunction.
According to the source, if the injunction is granted, Apple could gain access to OpenAI devices, storage drives and accounts.
That access could reveal more information about communications, file transfers and internal handling of the matter.
It could also expose unrelated internal material, creating additional reputational or legal complications.
OpenAI’s decision to publish selected messages may therefore be intended to build support before more extensive evidence becomes available through the court process.
Whether that approach helps will depend on what later discovery reveals.
The Dispute Reflects Wider Tensions Around AI Talent
Although the source focuses on the immediate lawsuit, the case also highlights the sensitivity surrounding employee movement between major technology companies.
Engineers often carry general knowledge and experience when they change jobs. Companies seek to protect confidential designs, research and internal systems that cannot legally be transferred.
The difficult question is where ordinary professional knowledge ends and protected trade secret information begins.
That distinction is particularly important in artificial intelligence and hardware, where competition for specialized employees is intense.
The Liu dispute may therefore become relevant beyond Apple and OpenAI if it clarifies how companies should manage employee departures, file access and confidential material.
OpenAI Faces Other Potential Legal Questions
The source also points to another possible area of future legal exposure: responsibility for autonomous actions carried out by AI agents.
That issue is separate from Apple’s lawsuit, but it reinforces the broader legal pressure surrounding OpenAI’s products and business practices.
As AI systems become more capable of acting with limited human supervision, companies may face questions about who is responsible when those systems access data, exploit vulnerabilities or perform unauthorized actions.
The Apple dispute concerns alleged human conduct and trade secrets. Future cases involving autonomous agents could create even more complex questions about intent, control and liability.
The court must decide whether Apple has shown enough evidence and urgency to justify a preliminary injunction.
The wider case will likely examine what Liu downloaded, whether he had authorization, whether OpenAI received the material and whether any information was used.
The published messages may become part of the factual background, but they are unlikely to determine the dispute alone.
Apple may also respond by providing additional context or arguing that OpenAI’s disclosures are irrelevant to the core trade secret claims.
The most important evidence will be the technical and documentary record rather than the companies’ public statements.
OpenAI has escalated its response to Apple’s trade secrets lawsuit by publishing emails and iMessages that portray Apple’s legal outreach as careless and Liu’s conduct as less secretive than alleged.
The materials may raise legitimate questions about Apple’s communication process and its description of events.
They do not directly resolve whether trade secrets were taken, transferred or used.
OpenAI’s disclosures may help it challenge Apple’s credibility, but they do not answer the central trade secrets allegation. The strategy could influence public opinion, while the outcome will ultimately depend on evidence about the files, access permissions and OpenAI’s involvement.
