Apple has asked the court overseeing its lawsuit against OpenAI and other defendants to allow its own experts to examine forensic images used in the case, while also requesting documents related to parts of OpenAI's hardware research and development.
See also: Apple sends legal letters to former employees who now work at OpenAI
In early July, Apple filed a lawsuit against former employees of OpenAI and io Products, accusing them of misappropriating trade secrets. The lawsuit involves former Apple employees who allegedly took confidential information with them when they left the company. Apple also alleges that former employee Chang Liu instructed another Apple employee on materials to review before an interview at OpenAI, while Tang Tan directed Apple job candidates to bring hardware components and other materials to interviews.
More recently, Apple said that an initial forensic inspection of a MacBook used by Liu after he left the company showed that he had used a confidential Apple circuit to run a simulation while working at OpenAI. Apple also noted that the MacBook indicated Liu's use of a Mac mini, which the company has since requested access to as part of the case.
OpenAI has refuted Apple's accusations, claiming that the situation is "a mess of Apple's own making," as the company's allegedly lax exit and security practices made it easier for employees to inadvertently retain access to confidential information after they left.
One of the points of contention in the case is Apple's request for expedited discovery, which OpenAI has opposed. Another is Apple's request for a temporary injunction that would restrict OpenAI's use of Apple's alleged trade secrets while the lawsuit progresses, potentially affecting its ongoing hardware development.
Apple argues that it needs expedited access to certain elements because OpenAI is still developing its hardware product, meaning that any alleged use of Apple's trade secrets could become further intertwined with that work and increasingly difficult to undo. OpenAI argues that Apple has not shown why a separate expedited discovery process is required and claims that Apple's requests are overly broad.
See also: RubyGems: OpenAI Agents Attacked the Platform
Because of this impasse, the court asked both sides to file a joint filing outlining their respective positions on expedited discovery. That document, filed last Friday, has now been made public.
In the joint filing, Apple is asking the court to allow its own forensic experts to examine the underlying forensic images that the defendants’ experts relied on to analyze Liu’s devices and determine whether he accessed or used Apple’s confidential information after he left the company. Apple claims that one of its forensic experts has already identified what it describes as “critical errors” in that analysis and argues that its experts need immediate access to determine whether there were other problems with the defendants’ forensic analysis.
Apple's forensic expert has uncovered several critical errors in the analysis performed by the defendants' experts, including failures to identify files on Mr. Liu's Mac Mini that were proven to have been used and stored on that device and incorrect conclusions that the presence of files with zero bytes or an empty kMDItemLastUsedDate on the Mac Mini means that those files were not accessed.
Apple argues that its experts' direct access to these forensic images is critical to understanding other errors that the defendants' experts may have made.
OpenAI argues that Apple's experts should not have unfettered access to the forensic images, suggesting instead that they be handled by a neutral forensic expert under a mutually agreed-upon protocol. Apple argues that this would put it at a disadvantage, as the defendants' own forensic experts already have direct access to the same material.
See also: Tata Electronics leak claims to expose Apple and Tesla trade secrets
In addition to forensic material, Apple is requesting expedited access to documents related to OpenAI's hardware development, including work on metal surfaces, power management, and battery functionality.
