Apple is one of the biggest companies of all time, but according to OpenAI this lawsuit doesn’t reflect that reputation for attention to detail. What happened exactly, and why does OpenAI say Apple was wrong from the start? Here I summarize the essentials and the evidence OpenAI published for you.
What OpenAI alleges
OpenAI says Apple first claimed it had contacted the company in February and received no response. Then Apple admitted its outside counsel had emailed the wrong person after confusing two Asian last names, and only acknowledged the mistake after OpenAI pointed it out.
Apple also said it had spoken with OpenAI’s general counsel; OpenAI says that conversation never happened. More than that, Apple didn’t raise the claims now in the lawsuit during that first contact and — after saying they were “resolving any issues” — didn’t reach out again for five months before filing suit.
Messages and emails that complicate the story
OpenAI publishes excerpts of iMessage and emails that highlight several critical points:
-
Apple employees asked Chang Liu, whose last day at Apple was January 22, 2026, for help locating files and content. That suggests Apple had active contact with that person before alleging improper access.
-
An email from Apple’s outside lawyer, Gabriel Gross, was sent by mistake to OpenAI’s legal team instead of the former employee. In that message Gross said he had spoken by phone, but OpenAI confirmed no such call took place.
-
Gross later clarifies the error and says Apple was “resolving any issues,” without having raised the formal accusations that now form the basis of the lawsuit.
About the allegations of residual access and trade secrets
Apple accuses Chang Liu of accessing confidential information after leaving the company. OpenAI responds that Apple itself asked Chang for help locating information, and that the so‑called residual access is a recurring issue in access management when someone departs Apple.
In the case of Tang Tan, Apple points to attempts to obtain or use trade secrets. OpenAI reminds readers that Tang worked at Apple for over 24 years and that the OpenAI team has repeatedly said they do not want or should not use third‑party confidential information.
Offer to talk and rejection of the preliminary injunction
OpenAI says it would have preferred that Apple raise these concerns before suing, and that it offered to cooperate to clarify everything. They argue that Apple’s request for a preliminary injunction is based on incorrect information and is unnecessary because OpenAI neither has nor wants Apple’s secrets. OpenAI insists its focus is building products and innovative technologies.
What this implies for the industry and for you
Sound familiar — the “residual access” problem? In large companies with thousands of employees it can happen: accounts, permissions and files remain active due to process failures. That doesn’t absolve anyone, but it does change the context for how these accusations should be investigated.
If you work in tech or lead teams, this is a reminder of three practical things you can do:
- Review and automate access revocation when an employee leaves.
- Keep clear traceability of internal requests for files.
- Communicate and document legal conversations before going to court.
In the end, the dispute raises more questions about internal processes and how big companies handle incidents between them. It’s not just a clash of two giants; it’s a reminder that human and procedural errors can escalate into litigation that affects reputation and resources.