Judge Opens OpenAI’s Hiring Files

A federal judge has ordered OpenAI to hand over hiring materials that Apple says could show whether its former employees carried confidential information into job interviews.

The ruling narrows the fight, which matters. Apple did not win a broad look into OpenAI’s hardware research. It did win expedited access to a smaller, more pointed set of records: interview checklists, scripts, guides, technical presentations and other materials linked to candidates from Apple who applied to OpenAI’s consumer hardware and robotics groups.

The court gave OpenAI five days to produce materials it has already collected and 14 days for other documents covered by the request, according to TechCrunch. Judge Edward Davila also allowed Apple access to certain private LINE messages involving people named in the case and approved depositions of former Apple employees Chang Liu and Alyssa Peng, as well as OpenAI chief hardware officer Tang Yew Tan.

That is a meaningful procedural win for Apple, though not proof of theft. Discovery orders decide what evidence can be examined, not what the evidence will show. Apple has accused OpenAI of a pattern of misuse of trade secrets, including claims that former employees downloaded confidential files from Apple laptops and coached others on what to bring to interviews. Those remain allegations within the lawsuit.

The judge rejected several of Apple’s wider requests, including expedited access to OpenAI research and development in areas such as metal finishes and power management, and broader demands for information on what former Apple employees built after joining OpenAI.

So the court’s message was selective rather than sweeping: Apple gets the interview trail first, not a guided tour of the lab. In trade secrets cases, that distinction is usually where the denominator lives.