A 2,000-word essay with chat log screenshots — no cease-and-desist warnings, just direct public rebuttal.

On the evening of August 3, OpenAI published an article titled "Apple Is Getting This Wrong" on its official blog, formally responding to the trade secret lawsuit Apple filed in late July. This isn't the tech industry's standard "no comment" or "we will respond in court." OpenAI chose an extraordinarily rare approach: laying the evidence directly on the table. The article not only refutes Apple's claims point by point but also attaches employee chat logs as supporting evidence, laying bare for the public how the two companies went from partnership to breakup.

Only a week had passed since Apple filed suit in California court. A week ago, most observers assumed OpenAI would choose silence — after all, when fighting a company like Apple, public confrontation is rarely the optimal move. But Sam Altman chose the most aggressive route: not just fighting the lawsuit, but winning the narrative battle first. The post quickly spread through tech circles after publication, topping Hacker News hot list and driving massive discussion on Twitter/X.

Why Public Rebuttal Instead of Private Settlement

To understand the weight of this, you need to know how tech company lawsuits normally play out.

The standard trade-secret lawsuit script goes like this: Company A sues, Company B issues a bland statement saying "we believe the claims are without merit," then both sides enter a long discovery phase, fight for a year or two, and eventually either settle or drop the case. The public almost never sees complete evidence — every exchange happens under sealed court filings. This approach works best for big companies: it satisfies legal department KPIs without making either side look too bad in public.

OpenAI threw this playbook out the window. The opening line sets the tone: "Apple's claims against us are wrong, and we want to set the record straight." Then point-by-point rebuttal: on the claim that a departing employee took trade secrets, OpenAI releases the employee's chat logs with his Apple manager before departure, showing Apple explicitly saying "you don't need to bring anything over"; on claims that OpenAI accessed sensitive information during the Siri integration, OpenAI details the access agreements both sides signed, proving all data access stayed within Apple-drawn boundaries; on claims of "improper use" of Apple technology, OpenAI posts emails from Apple engineers proactively sending technical suggestions.

It's a ruthlessly effective move. The core of trade secret litigation is information asymmetry — the plaintiff can vaguely claim "trade secrets were stolen" without producing concrete evidence upfront. But by dropping chat logs publicly, OpenAI kicks the ball back to Apple's court: you say I stole something, I'm releasing the conversations — where exactly is the theft? If Apple can't produce stronger evidence, they've already lost half the battle in the court of public opinion.

Honeymoon to Breakup: The Interest Conflicts Behind Siri Integration

To understand the root of this lawsuit, go back to WWDC 2025.

That's when Apple took the stage to announce its partnership with OpenAI — ChatGPT would be deeply integrated into Siri in iOS 18. Tim Cook and Sam Altman smiled side by side, and media called it a "deal of the century." Apple got the strongest LLM capabilities available, finally able to deliver in the AI era; OpenAI got access to a billion-plus iPhone users. It looked like a win-win.

The honeymoon was shorter than anyone expected. Two fundamental conflicts derailed it. First, data control. Under the partnership agreement, details around how user data from Siri-ChatGPT calls would be used, stored, and what permissions each side had were never fully resolved from the start. Apple insisted user data must stay on Apple's side; OpenAI argued reasonable model improvement required access to interaction data. The two sides haggled for months, and the integration experience was never as smooth as the onstage demo suggested.

Second, product roadmap conflicts. When OpenAI accelerated its desktop app and Agent ecosystem push in 2026, it directly touched Apple's bottom line — Apple doesn't allow an uncontrolled AI entry point on its platform. When OpenAI released ChatGPT Desktop's computer control feature, Apple repeatedly demanded changes citing security concerns, and the conflict went fully public.

In Issue 026 we covered Apple's lawsuit against OpenAI, judging it as "tactical pressure" to get better terms in partnership renewal negotiations. But OpenAI's forceful public response suggests things may be more serious than we thought — OpenAI seems to have abandoned maintaining surface-level peace with Apple and is ready for a bare-knuckle fight.

The Rules of War Have Changed in the AI Era

The most notable thing about this lawsuit isn't who's right or wrong — it's the shift in warfare rules it foreshadows.

In the PC era and mobile internet era, tech giant wars played out on three main battlefields: product launches, app store review, and court. The first two had PR teams and review rules as buffers; the third had NDAs and sealed filings as fig leaves. The public always saw carefully packaged narratives, with the real kill shots happening under the table.

But OpenAI chose a fourth battlefield: evidence warfare in the public square. Releasing raw chat logs directly, letting the public judge who's right and wrong. Few companies used this tactic before the AI era, because tech dispute details were usually too technical for the public to understand or care about. But AI is different — ChatGPT and Siri are products used by hundreds of millions daily, and the outcome of this lawsuit directly affects user data security and product experience. The public has strong motivation to watch.

More importantly, as an AI company whose stated values include "transparency," the evidence-release play aligns perfectly with OpenAI's brand. When your opponent is a consumer electronics company famous for its culture of secrecy, "I'm putting everything out there for everyone to see" is itself a devastating PR attack. You say I stole from you, Apple? Fine — I'm releasing all the chat logs. Do you dare do the same?

This may set a precedent. Future tech giant lawsuits will see more companies choosing to make key evidence public. Rather than letting a judge decide in a closed courtroom, why not let hundreds of millions of netizens be the jury first? In the AI era, public trust itself is the core commercial asset.

Who's the Real Loser

Whoever wins the lawsuit, one thing is virtually certain: the deep partnership between Apple and OpenAI is effectively dead.

For Apple, this means Siri's AI upgrade path gets rockier. Capabilities originally plugged by ChatGPT will now have to be built slowly with in-house models, or replaced through new deals with Google or Anthropic. But Google owns Android; Anthropic is closer to Amazon. Neither option is perfect. We covered in Issue 027 how Siri's revamp landed with a thud — without strong third-party model backing, the Apple Intelligence experience gap will only widen.

For OpenAI, losing the iPhone super-entry-point is a loss, but far from fatal. ChatGPT's own app has hundreds of millions of MAUs, desktop and Agent ecosystems are growing fast, and enterprise revenue is surging. More importantly, OpenAI is proving one thing: no single platform can put a chokehold on me. That signal to other platform holders (Google, Microsoft, device makers) carries more deterrent weight than the cost of losing one lawsuit.

What's truly worth watching is the impact on the whole AI ecosystem. When two giants go from partnership to confrontation, the real losers are users — the Siri on your phone and the ChatGPT you use may never connect smoothly again. The walled garden isn't being torn down in the AI era; it's being built higher.

But then again, this might be the nature of tech. No permanent friends, no permanent enemies — only permanent interests. Shaking hands and smiling at a keynote one day, dumping evidence in a blog post the next. Rather than hypocritical peace, maybe we should appreciate this kind of straightforward clash — at least you know what they're fighting over, at least the evidence is out there for you to judge.

See you tomorrow.