Apple Just Sued OpenAI. It’s Really a War Over the Next iPhone.

Jul 13, 2026 | apple ai

Two years ago, Tim Cook and Sam Altman shared a stage to announce that ChatGPT would live inside the iPhone. On Friday, Apple sued OpenAI for trade secret theft. That arc — from partnership to lawsuit in twenty-four months — is the story of the AI era in miniature. But underneath the courtroom drama is a bigger question that matters far more to Europe than who wins: who builds the device that comes after the smartphone. gafam.ai reads both layers.

What Apple Alleges — Stated as Allegations

Everything that follows is Apple's account, drawn from a complaint it filed on July 10 in the US District Court for the Northern District of California. These are allegations, contested and unproven.

Apple accuses OpenAI, the Jony Ive-founded hardware startup io Products (which OpenAI acquired), and two former Apple engineers of a coordinated scheme to steal Apple's trade secrets to build OpenAI's consumer hardware. In its filing, Apple does not hedge: it claims that at every level, from members of its Technical Staff to its Chief Hardware Officer, OpenAI has been stealing Apple's trade secrets and confidential information.

The central figure is Tang Tan — OpenAI's Chief Hardware Officer, and before that a 24-year Apple veteran who rose to VP of product design for the iPhone and Apple Watch. Apple alleges Tan used confidential Apple project codenames during OpenAI's recruiting to extract more information from Apple employees interviewing for jobs; asked those still-employed Apple candidates to bring actual parts — batteries, logic boards — to interviews for show and tell sessions; and circulated an internal Apple offboarding document to teach new hires how to evade Apple's exit-security checks. The second named engineer, Chang Liu, allegedly failed to return an Apple laptop and downloaded confidential technical documents after leaving.

Apple goes further, alleging OpenAI approached Apple's own trusted suppliers with confidential information — even, it claims, having one partner perform a trade-secret metal-finishing technique under the false impression it had Apple's permission. Apple says over 400 of its former employees now work at OpenAI, and that it wrote to OpenAI in February and received no response. It is seeking an injunction, the return of materials, and damages.

OpenAI's Response — Given Fairly

OpenAI rejects the accusations. In a statement, the company said it has no interest in other companies' trade secrets and remains focused on building innovative technology that empowers people everywhere. Altman, posting on X, struck a notably respectful tone toward Apple — saying he is not afraid of Apple but has tremendous respect for them, calling it an s-tier company — a deliberate contrast to his combative posture in other recent disputes.

Two points of context are essential for fairness. First, this is a one-sided complaint at the opening of litigation; Apple's dramatic framing (it calls the allegations the tip of the iceberg) is advocacy, not proven fact. Second, the talent flow at the heart of the case is partly a structural feature of California law, which bars non-compete agreements and makes exactly this kind of employee movement between rivals legal and common. The line between a departing engineer's own knowledge and a company's protectable trade secret is genuinely contested terrain — which is what courts exist to resolve.

Why This Is Really About Hardware

Strip away the accusations and the motive is clear: OpenAI is building a device, and Apple is defending the crown jewel that trade secrets protect — not software, but hardware and industrial design.

OpenAI has spent the past year assembling a hardware capability, most visibly by acquiring Jony Ive's io Products for roughly $6.4 billion, with Ive now leading its device work. Altman has said prototypes are finished. Pair that with OpenAI's move into custom silicon (its Jalapeño inference chip) and its models, and the shape of the ambition is unmistakable: OpenAI is assembling the same vertically integrated stack — silicon, device, and operating intelligence — that made Apple the most valuable company on earth. It is trying to build the iPhone of the AI age. And the people who know how to build an iPhone work, or worked, at Apple.

That is why this suit matters beyond the two companies. The next general-purpose computing device — the thing that might do to the smartphone what the smartphone did to the PC — is being contested right now. And every party to that contest is American: Apple defending, OpenAI and Ive attacking, with Meta's smart glasses, Musk's rumoured device and Google in the wings.

The European Perspective

For Europe, the Apple-OpenAI lawsuit is a window onto a race in which Europe is not a competitor, not a spectator with a stake, but simply absent. The fight over the next computing device — the post-smartphone AI hardware platform — is being waged entirely among American companies, and its outcome will shape the layer through which the next generation touches technology as profoundly as the iPhone shaped the last. Europe has no entrant.

There is no European Apple defending an integrated hardware empire, no European OpenAI-and-Ive attempting to build the successor, no European smart-glasses or AI-device contender at scale. Whatever emerges from this contest — an OpenAI device, a defended Apple ecosystem, or a Meta wearable — Europe will once again be the market, not the maker.

This connects to the deeper pattern gafam.ai has documented all summer, but adds a distinct and underappreciated dimension: capability transfers not only through models, as in the distillation disputes we covered, but through people. Four hundred former Apple employees now at OpenAI is the human version of the same anxiety that drove Meta to wall off Claude Code and Anthropic to accuse Alibaba — the fear that your hardest-won advantage walks out the door. And here Europe faces a structural asymmetry it rarely discusses: the fluid talent mobility that California's non-compete ban enables is precisely what lets American firms recombine expertise at speed, for better and worse.

Europe's more restrictive labour and non-compete regimes protect incumbents but also slow exactly the recombination that produces new hardware entrants. The lesson is not that Europe should copy California wholesale, but that it should recognise the connection between talent fluidity and the emergence of hardware challengers — and ask why no European contender for the next device exists. Regulation can govern how the winning device treats European citizens, as the Meta likeness case showed. It cannot conjure a European device into being.

The next platform is being fought over now, in an American courtroom, by American companies, and Europe's absence from that fight is the quietest but most consequential fact in the story. gafam.ai will be watching.

We are not first. We are right.

🔒 This analysis is for GAFAM Intelligence members only.

→ Become a Member

Already a member? Log in here