OpenAI Takes Apple’s Lawsuit to the Court of Public Opinion — and Brings Receipts

Apple’s legal team probably didn’t expect their opponent to publish a blog post at midnight. But that’s exactly what OpenAI did, dropping a scathing response titled “Apple is getting this wrong” that reads less like a legal brief and more like a subtweet with attachments. The ChatGPT maker didn’t just deny stealing trade secrets — they called the lawsuit “careless, aggressive, and oddly personal,” then backed it up with iMessage screenshots and email threads that make Apple’s claims look shaky at best.

The PR Offensive

This wasn’t a court filing. It was a press play. OpenAI knows the actual litigation will drag on for years, but the narrative war happens now. By publishing private communications — something companies usually guard like nuclear codes — they’re betting that transparency looks better than silence. The messages show Apple executives discussing OpenAI’s public research papers, not proprietary data. One exchange even has an Apple researcher asking for clarification on a published method. That’s not theft. That’s literature review.

Apple’s complaint centers on a former employee, Peng Xiao, who allegedly downloaded thousands of files before jumping to OpenAI. The suit paints Xiao as a corporate spy. OpenAI’s counter-narrative: Xiao was a researcher who saved his own work, much of which overlapped with public domain knowledge. The screenshots they released show Apple’s own security team initially clearing Xiao’s departure. The “aggressive” label starts to stick.

Why This Feels Personal

“Oddly personal” is the phrase doing heavy lifting here. Apple doesn’t typically sue AI labs over talent poaching — they acquire them. Remember Turi? Laserlike? Xnor? Apple buys the team, folds the tech into Silicon, and moves on. Suing a former employee and his new employer suggests something else: fear. Apple’s AI strategy has been visibly flat-footed. Siri still can’t hold a contextual conversation. Their on-device models lag behind what Google and Microsoft ship. Meanwhile, OpenAI just dropped GPT-4o with voice mode that feels like the assistant Apple promised a decade ago.

The lawsuit reads like a company trying to slow down a competitor while it catches up. Discovery will be brutal for Apple. Every internal email about “catching up to ChatGPT” becomes evidence that this isn’t about trade secrets — it’s about market position.

The Talent War’s New Rules

Tech’s unwritten rule used to be: people move, knowledge moves with them, nobody sues unless code gets copied verbatim. Apple just broke that norm. If this stands, every researcher who switches jobs becomes a litigation risk. That chills hiring across the industry. OpenAI’s public response signals they won’t play defense. They’re framing this as an attack on open research — and they’re not wrong. The methods in question? Published. The architectures? Documented in conference papers. The “secrets” appear to be engineering intuition, not IP.

Xiao’s LinkedIn shows a PhD in machine learning from Carnegie Mellon. He published at NeurIPS. His “trade secrets” are his expertise. Courts have historically protected an employee’s right to use general knowledge and skill. Apple’s theory would effectively indenture researchers.

What Happens Next

Apple will file a formal response. Their lawyers will argue the blog post prejudices the jury pool. OpenAI’s lawyers will say it corrects a false public record. The judge will probably hate both of them. But the damage to Apple’s reputation among AI talent is done. Researchers talk. They read Hacker News. They see a company suing a peer for taking his brain to a better offer. The best ones won’t interview at Cupertino. They’ll go where the work ships.

OpenAI’s midnight blog post was a risk. Publishing private messages waives privilege. But it also forced Apple to litigate in daylight. That’s where the truth usually wins.

Source: Jess Weatherbed