Apple Gets Access to OpenAI Interview Records in Trade-Secrets Case

Apple has won a court order giving it expedited access to selected OpenAI hiring records, employee communications and testimony as part of a lawsuit alleging that confidential Apple information was improperly obtained and used. The ruling, issued by US District Judge Edward Davila in California, covers records connected to former Apple employees who interviewed for positions in OpenAI’s consumer hardware and robotics operations.

The decision gives Apple a new route to investigate what was discussed during recruitment and how former employees handled potentially confidential information after leaving the iPhone maker. OpenAI has disputed Apple’s allegations, and the judge has not determined that trade secrets were stolen. The court also rejected or deferred several broader discovery requests, meaning Apple’s access to OpenAI’s technical work remains limited in important areas.

Key Takeaways

• Apple won expedited access to selected OpenAI hiring records, including interview guides, checklists, scripts, instructions and technical presentations.
• OpenAI must produce already collected materials within five days and other requested materials within 14 days, according to reporting on the ruling.
• The judge approved depositions of former Apple employees Chang Liu, Alyssa Peng and Tang Yew Tan, who is now OpenAI’s chief hardware officer.
• Apple was also granted access to specified private LINE messages and other evidence relevant to its allegations.
• The court did not grant every request: some technical research and development records, including material concerning metal finishes and power management, were not ordered produced on an expedited basis.
• The ruling concerns evidence gathering. It does not establish that OpenAI or the former employees are liable for trade-secret misappropriation.

What Did the Court Order in Apple’s Lawsuit Against OpenAI?

The ruling gives Apple faster access to evidence that could help it investigate its claims about OpenAI’s recruitment of former Apple staff. The requested material includes documents used to guide or conduct interviews for positions in OpenAI’s Consumer Hardware Products Group, along with records summarising discussions with candidates.

According to TechCrunch’s reporting, the categories include interview checklists, guides, instructions, manuals and scripts, as well as technical presentations or records referring to Apple technology and components. The order also covers certain forensic evidence and communications relevant to the dispute.

The court set separate deadlines for production. OpenAI was given five days to provide materials it had already collected and 14 days to produce other materials covered by the request. These deadlines accelerate the exchange of specified evidence rather than requiring OpenAI to disclose all its internal research or every document relating to former Apple employees.

The decision is important because the case involves information that could be difficult to reconstruct later. Apple has argued that evidence on devices and in communications could help establish whether confidential material was accessed, retained, discussed or used after employees departed.

Bloomberg Law reported that the judge identified a risk that important forensic evidence could be lost, citing a conversation between former Apple employees about wiping Apple-owned electronics. That concern was part of the rationale for allowing certain discovery to proceed quickly. It is not, by itself, a final determination that evidence was improperly destroyed.

Why Is Apple Suing OpenAI?

Apple filed its lawsuit in July 2026, accusing OpenAI and former Apple employees of misappropriating trade secrets and confidential company information. The allegations concern the movement of employees from Apple to OpenAI and the possibility that sensitive engineering material was carried into the AI company’s hardware development work.

Apple has alleged that confidential information included technical specifications, engineering presentations, project information and details about products that had not been announced publicly. It has also alleged that recruitment discussions involved questions about Apple technologies and components.

A central issue is whether the information at the centre of the dispute was genuinely confidential, whether it was accessed or disclosed without authorisation, and whether any such conduct can be connected to OpenAI or particular employees. Those questions are not settled by the latest discovery ruling.

The case also reflects the growing strategic importance of hardware for AI companies. OpenAI is developing consumer hardware ambitions of its own, while Apple has an established ecosystem of devices, operating systems and services. If a dispute involves confidential product designs, engineering processes or unreleased technology, the potential commercial consequences could extend beyond a single hiring decision.

However, competition between the companies does not itself prove wrongdoing. The court must assess the evidence and the legal claims rather than infer misconduct from the fact that employees moved between competing businesses.

What Interview Records Can Reveal

The interview documents could help establish what OpenAI asked former Apple employees during recruitment and what candidates were instructed to bring or discuss.

Interview guides and scripts: These may show whether interviewers followed a standard process, what subjects they were told to cover and whether the process included questions about technology developed at a previous employer.

Technical presentations and interview summaries: These could help reconstruct what candidates said about engineering problems, components or product designs. The critical issue is whether any discussion crossed the line from general professional experience into protected confidential information.

Internal instructions and manuals: Such documents may indicate what OpenAI told interviewers and candidates about handling information belonging to other companies. They could also provide context for claims that candidates were expected to follow boundaries around confidential material.

Forensic records: Images of devices and other forensic evidence may help investigators understand which files were present, whether certain files were accessed and when relevant activity occurred. Interpretation can be complicated: the presence of a file on a device does not automatically prove that it was deliberately taken, shared or used to develop a competing product.

Taken together, the records may help the parties test their competing accounts. They could support parts of Apple’s allegations, support OpenAI’s response or show that particular events have been characterised incorrectly by either side.

Three Former Apple Employees Face Depositions

The judge also approved depositions of Chang Liu, Alyssa Peng and Tang Yew Tan. Depositions are formal, out-of-court sessions in which witnesses answer questions under oath. Their testimony may help clarify their work at Apple, their departure from the company, their recruitment by OpenAI and their knowledge of the issues raised in the lawsuit.

Liu and Peng are former Apple employees who joined OpenAI. Apple’s allegations include claims involving communications between former colleagues and the handling of confidential information. The court’s approval of access to specified LINE messages could allow Apple to review relevant conversations in greater context.

Tan, OpenAI’s chief hardware officer, is particularly important because Apple has alleged that recruitment involving the hardware organisation raised questions about confidential Apple technology. Apple has accused him of asking candidates about unreleased products and encouraging discussions or demonstrations involving components from their previous work.

These remain allegations, not established facts. OpenAI and the individuals named in the case have disputed Apple’s characterisation of the events. The depositions are intended to gather evidence, and their significance will depend on what the witnesses say and how their testimony compares with documents and other evidence.

The case also illustrates why context matters in employee-mobility disputes. Engineers and hardware specialists routinely bring general skills, experience and professional knowledge from one job to another. Trade-secret law, by contrast, protects qualifying confidential information against certain forms of improper acquisition, use or disclosure. The legal question is not simply whether a former employee discussed a previous employer’s work.

The Judge Rejected Some of Apple’s Broader Requests

Although Apple secured faster access to several categories of evidence, it did not receive everything it requested.

The court declined to expedite the production of certain OpenAI research and development materials concerning areas such as metal finishes and power management. It also rejected several broader requests involving OpenAI’s hardware research and work performed by former Apple employees after they joined the company. Reporting indicates that some requests may be pursued again if Apple narrows their scope.

This distinction is important. Interview records and communications can be directly relevant to questions about recruitment, what candidates discussed and what instructions were given. A demand for a wider range of technical development records raises additional questions about relevance, burden and the risk of exposing a company’s own confidential information.

Apple’s position is that some OpenAI development work may be connected to the alleged use of Apple-originated information. OpenAI has argued that broad demands could expose its research and development unnecessarily and extend beyond what is needed to resolve Apple’s claims.

The court’s partial decision reflects the need to balance access to evidence against those concerns. The fact that some requests were denied or delayed does not mean Apple’s overall lawsuit has failed. Equally, the approval of expedited discovery does not mean the court has accepted Apple’s allegations as true.

Why the Dispute Matters for AI Hardware

The lawsuit arrives as AI companies seek ways to bring their technology into new devices and consumer experiences. Hardware development can involve years of work on industrial design, materials, power efficiency, component selection, thermal performance and manufacturing. Confidential information in these areas may have commercial value before a product is announced.

For Apple, the case is about protecting information it says belongs to the company and preventing competitors from benefiting from confidential work. For OpenAI, the dispute concerns how it recruits experienced engineers and develops products while denying allegations that it sought or used Apple’s trade secrets improperly.

The outcome could have implications for recruiting practices across the technology industry. Companies hiring engineers from competitors may face increased scrutiny over interview questions, technical demonstrations, files brought to interviews and the treatment of information from previous employers.

Businesses can reduce legal risk by documenting clear rules: candidates should not bring confidential files or proprietary components from a previous employer; interviewers should avoid questions seeking protected information; and new hires should have defined processes for identifying and reporting potentially sensitive material.

Such safeguards do not resolve the factual disputes in this case, but they show why hiring policies and information-security controls matter when employees move between competing technology companies.

What Happens Next?

The immediate next step is the production and review of evidence covered by the court’s order, followed by depositions and further litigation over disputed discovery and the merits of the claims. The records could sharpen the parties’ arguments about what was discussed during recruitment and whether any confidential Apple information was improperly handled.

The broader case remains unresolved. Apple will need to establish the elements of its claims, while OpenAI and the other defendants can challenge the allegations and the interpretation of the evidence. Further court decisions will determine whether the lawsuit proceeds, what additional material must be disclosed and whether any restrictions on the use of disputed information are warranted.

Frequently Asked Questions

1. Why did Apple get access to OpenAI interview records?

A US federal judge granted Apple expedited access to selected hiring records and other evidence relevant to its allegations that confidential Apple information may have been misappropriated. The records may help establish what candidates were asked and whether Apple technology was discussed.

2. Has the court ruled that OpenAI stole Apple trade secrets?

No. The ruling concerns evidence gathering, not a final decision on liability. Apple’s allegations remain contested, and the court has not established that OpenAI stole Apple’s trade secrets.

3. Which former Apple employees are involved in the depositions?

The judge approved depositions of Chang Liu, Alyssa Peng and Tang Yew Tan, OpenAI’s chief hardware officer. Their testimony may address recruitment, their work history and issues raised by the lawsuit.

4. Did Apple get access to all of OpenAI’s hardware research?

No. The judge declined to expedite some broader technical discovery requests, including material related to metal finishes and power management. Some requests were rejected or deferred, and the scope of any further disclosure remains subject to the court’s decisions.