Apple Sued OpenAI Over Stolen Secrets. The Evidence Is Two Former Apple Employees.

Apple filed suit against OpenAI in federal court in Northern California on 10 July 2026, alleging trade secret theft connected to OpenAI's push into consumer hardware. The filing accuses OpenAI's hardware chief, Tang Tan - a former Apple vice president of product design - and Chang Liu, a former Apple engineer, of a coordinated pattern of misconduct that Apple's lawyers characterised as occurring "at every level" of the interaction between the two companies.

The specific allegations

According to Apple's filing, Tan directed job candidates who were still employed at Apple to bring "actual parts" from Apple's supply chain to their OpenAI interviews, for what the company describes as "show and tell" sessions designed to elicit further confidential information from prospective hires who had not yet formally left their positions. Separately, Apple alleges that Chang Liu stole a company laptop on his way out the door, and that OpenAI more broadly coached departing Apple employees on how to evade the company's internal security and exit processes.

OpenAI's public response has been a flat denial: "We have no interest in other companies' trade secrets." Sam Altman addressed the suit directly on X, writing that he was "not afraid of Apple" and calling the company an "s-tier" business - a notably respectful tone given the seriousness of the underlying allegations. Elon Musk then inserted himself into the public conversation, trading accusations of "scamming" with Altman in a separate, parallel exchange that generated its own wave of coverage, layering a personal rivalry on top of the corporate dispute.

Perhaps the most striking number in the entire filing is contextual rather than allegation-specific: more than 400 former Apple employees are now reported to work at OpenAI. Whatever the outcome of the specific trade secrets claims, that figure alone describes a substantial, sustained flow of talent and institutional knowledge from one company to a direct emerging competitor in hardware.

Why this case is genuinely unusual - and why it isn't

Talent poaching between competitors, including poaching that raises trade secret concerns, is not a new phenomenon in the technology industry - it has generated litigation for decades, from the "no-poach" agreements Silicon Valley giants were sued over in the 2010s to countless smaller, quieter disputes over departing engineers and the code, designs, or client lists they took with them.

What makes this specific case notable is less the underlying dynamic and more the profile of the two companies involved: Apple and OpenAI, both instantly recognisable well outside the technology industry, engaged in a dispute over the movement of people and knowledge between them, playing out in public with commentary from figures as prominent as Musk. The mechanics being alleged - a hiring manager encouraging candidates to bring confidential materials to an interview, a departing employee removing company property - are mundane by the standards of corporate espionage litigation. They simply happen, in this instance, to involve two of the most closely watched companies on earth.

The underlying question every hiring business should sit with

Strip away the celebrity of the two companies involved and the case reduces to a question that applies, in some form, to almost every hire made from a competitor, at any scale: when someone joins a new company from a rival, what exactly comes with them - and whose job was it to check?

Most of what an experienced hire brings from a previous employer is, straightforwardly, theirs to bring: skills, general industry knowledge, professional relationships built over a career, and judgment developed through experience. None of that is remotely controversial, and restricting it would make normal labour mobility impossible.

A narrower category is genuinely not theirs to bring: specific confidential information belonging to the former employer - supplier terms, unreleased product specifications, proprietary processes, client data protected by contractual confidentiality obligations. This is the category Apple's suit alleges Tan and Liu crossed into, and it is the category that creates real legal exposure, for the individual and potentially for the hiring company, when it is crossed.

Why this matters beyond big tech

Recruitment agencies, in particular, sit in an unusually direct position relative to this risk, because sourcing candidates directly from a client's competitor is a routine, often explicitly requested part of the job. When an agency places a senior hire straight from a competitor's payroll into a new role, the standard due diligence typically covers track record, cultural fit, and reference checks on performance and conduct. It far less often includes any explicit check on what confidential information, systems knowledge, or client relationships that person might be bringing with them - and whether bringing it exposes either the candidate or the new employer to legal risk.

This gap is not usually a problem, because most hires do not involve genuinely sensitive confidential information, and most departing employees do not attempt to take anything they shouldn't. But the Apple-OpenAI case is a useful, high-visibility reminder that the gap exists, and that when a hire does go wrong in this specific way, the consequences - as this lawsuit demonstrates - can be severe, public, and expensive for everyone involved, including the hiring company that never asked the question.

A practical addition to placement due diligence

For recruitment agencies and hiring businesses more broadly, the practical response is not to treat every hire from a competitor with suspicion - that would be both impractical and unfair to the overwhelming majority of candidates who bring nothing improper with them. It is to add a specific, deliberate question to the placement process for senior or technically sensitive roles: does this candidate's move involve any material that is genuinely confidential to their previous employer, and has the candidate been made explicitly aware of their obligations around it?

Asking that question does not guarantee a case like Apple v. OpenAI never happens. It does mean the agency and the hiring company can demonstrate, if it ever comes to it, that the question was asked - which is considerably better legal and reputational footing than the alternative of having never asked at all.