Apple has expanded its trade-secret case against OpenAI after alleging that a former Apple engineer accessed confidential technical information while working for the artificial intelligence company and used proprietary material to train an AI system.
The allegation concerns Chang Liu, who previously worked as a senior system electrical engineer at Apple before joining OpenAI. Apple has alleged that Liu accessed a confidential power-converter circuit and later used information related to it to train an artificial intelligence agent.
The claims were disclosed in a filing before a US federal court during the exchange of evidence between the companies. Apple said it discovered the new information after examining a MacBook that OpenAI provided as part of the litigation.
Algoritha Security Launches ‘Make in India’ Cyber Lab for Educational Institutions
What are trade secrets, discovery and AI training?
A trade secret is confidential business information that gives a company a commercial advantage and is protected because the company takes steps to keep it secret. In a technology company, this can include engineering designs, source code, technical drawings or specialised manufacturing information.
Discovery is the stage of a lawsuit in which opposing sides exchange relevant documents, electronic devices, records and other evidence. It allows each side to investigate the claims and prepare its case before the court makes a final decision.
AI training is the process through which an artificial intelligence system learns patterns from data. Apple’s allegation is significant because it claims that confidential technical information was not merely viewed but was allegedly used as part of training an AI system.
The allegations have not been established as a final judicial finding. OpenAI has disputed the lawsuit and asked the court to dismiss Apple’s claims.
Apple says a MacBook revealed new evidence
According to Apple’s filing, the company’s latest allegations emerged after it examined a MacBook supplied by OpenAI on August 21.
Apple said its examination revealed information that raised questions about Liu’s access to confidential material after he joined OpenAI.
The company alleges that Liu accessed a confidential schematic connected to one of Apple’s power-converter circuits while working at OpenAI. Apple further claims that proprietary information from the circuit was used to train an artificial intelligence agent in March 2026.
Apple considers circuit designs and related engineering information to be proprietary. The alleged access and subsequent use therefore form an important part of the company’s wider trade-secret dispute with OpenAI.
The new evidence has also prompted Apple to seek faster discovery. The company wants access to additional documents, electronic devices and other material that could help establish what information was accessed, how it was accessed and what happened to it afterwards.
OpenAI wants the lawsuit dismissed
OpenAI has opposed Apple’s legal action and asked the court to dismiss the lawsuit.
The company has defended its development work and maintained that it is creating products that are new rather than simply reproducing another company’s technology.
The disagreement now involves two separate legal positions. Apple wants broader and faster access to evidence based on what it says it found on the MacBook, while OpenAI is challenging the lawsuit itself.
The court’s decisions on discovery could therefore become important. If Apple receives access to additional devices and records, investigators and lawyers may be able to establish a clearer timeline around Liu’s alleged access to the confidential material.
The case raises a new AI-era trade-secret question
The dispute goes beyond a conventional allegation that an employee took a confidential document.
Apple is alleging that technical information could have been used in the development or training of an AI system. That raises a more complicated question about how proprietary engineering information should be protected when companies increasingly use AI tools to analyse and process large amounts of data.
Hardware designs can contain commercially valuable information even when they are only one part of a larger product. A power-converter circuit, for example, can reveal engineering decisions about how a device manages and distributes electrical power.
The legal issue will ultimately depend on evidence. The court will have to consider whether the information involved qualifies for trade-secret protection, whether Liu accessed it improperly and whether Apple can establish that its proprietary information was subsequently used in the manner alleged.
For Apple, the immediate objective is to obtain more evidence. For OpenAI, the immediate legal strategy is to have the lawsuit dismissed.
The case is therefore moving into a potentially important stage in which the handling of electronic evidence could determine how far Apple’s allegations can proceed.
What this means for you: The case shows why confidential technical information can remain valuable even when it exists only inside company devices, documents or engineering systems. Employees working with sensitive corporate data should follow access restrictions and never copy or use proprietary material for another employer or outside project.
For technology companies, the dispute also highlights the need for clear controls over what information employees and AI systems can access. As AI becomes part of engineering and software workflows, protecting confidential data will increasingly depend on both traditional legal safeguards and technical access controls.
