Elon Musk has stumbled! While suing OpenAI, he secretly distilled ChatGPT.

Editor: Taozi

[New Intelligence Yuan Introduction] A major fiasco! Elon Musk, who sued OpenAI for "betraying its mission," actually admitted personally that his own Grok model distilled ChatGPT.

On the fourth day of the court battle, Musk dropped a bombshell!

On April 30, at the federal court in Oakland, California, the fourth day of the lawsuit filed by Musk against OpenAI began.

At the scene, William Savitt, the chief lawyer of OpenAI, stood up and asked a question that silenced the entire courtroom for three seconds:

Did xAI distill OpenAI's models?

Musk first evaded the question, saying, "All AI companies are doing this." Savitt continued to press, so the answer is yes?

Partly.

As soon as these words were spoken, the whole internet was stunned!

Musk admitted that the Grok model distilled ChatGPT

A person who is suing OpenAI for "betraying its non - profit mission" admitted that his own AI distilled the competitor's model.

A major double - standard scene, tearing up the pretense

It has to be said that the irony of this scene is overflowing the screen.

On the witness stand in the court, Musk's casual statement completely broke the ice -

This is the standard practice in the AI industry, and all AI companies are doing it.

Musk admitted that the Grok model distilled ChatGPT

Distillation, in essence, is to let the competitor's AI be a tutor and teach a student of similar level at a low cost.

Musk sued OpenAI, accusing Altman of violating the non - profit mission and using the $38 million he donated to build a for - profit company valued at $800 billion.

He is the plaintiff, the one who accuses others of "stealing charity."

But in the courtroom of this case, he admitted another kind of "stealing."

Is distillation illegal? Currently, it's a gray area in the law: it may not be illegal, but it almost certainly violates OpenAI's terms of service.

Silicon Valley giants are walking on the edge of copyright law when it comes to training data, scraping a large amount of internet content and claiming "fair use" when sued by copyright holders.

Anthropic claims that OpenAI violated the terms of service by using its API

Now even Musk says: everyone is doing it.

This answer was so subtle that it made the lawyer silent for a few seconds. He admitted, but not completely; he confessed, but left some leeway.

Anthropic first, xAI last?

There was also a detail during the court session that sparked a heated discussion. Musk personally ranked global AI companies.

When asked about his boast last summer that "xAI will soon surpass all companies except Google," Musk's ranking in court was:

Anthropic first, OpenAI second, Google third, and open - source models fourth.

Elon Musk

He said, "xAI is very small, only about one - tenth the size of OpenAI, with only a few hundred employees."

This is completely different from his usual style on X.

The man who shouts "Grok is on a rampage" every day on social media suddenly became extremely humble on the witness stand.

In court, Musk needed to describe xAI as small as possible to refute the accusation that "you sued OpenAI to attack your competitors."

OpenAI social media

To put it bluntly, to win the lawsuit, you have to admit that you're not good enough first.

Even harsher than Musk's testimony

The most memorable moment during the court session came from Judge Yvonne Gonzalez Rogers.

Musk's lawyer wanted the expert witness to discuss that AI might lead to "human extinction," but was stopped by the judge on the spot.

Musk himself also repeatedly mentioned the "Terminator" scenario in his testimony, saying that he wanted to ensure that Tesla's robots "don't become Terminators."

Movie screenshot

Who would have thought that the judge's response was a textbook example -

I've noticed that despite these risks, Mr. Musk himself is creating a company in "this field."

I believe many people are reluctant to entrust the future of humanity to Musk.

But that doesn't matter. This is a case about whether a charitable trust has been violated, not a trial about the safety risks of AI.

The subtext of this sentence is much sharper than its literal meaning.

It points to a fact that Musk has been trying to avoid: you're shouting that AI will destroy humanity, but at the same time, you've founded an AI company, distilled your competitor's model, and merged it with SpaceX.

Are you saving humanity or competing with humanity for business?

OpenAI President's diary, the real time - bomb

Musk's testimony is just the prologue of this trial.

Next Monday, Greg Brockman, the co - founder of OpenAI, will testify in court.

Brockman is a key witness because his private diary has been accepted as evidence by the court. There are several passages like this:

Sam Altman

His story will be: we were finally dishonest with him about still wanting to go for - profit, just without him.

It's wrong to steal the non - profit organization from him.

Converting to a B - corp without him is morally bankrupt, and he's not stupid.

These diary entries are one of the key reasons why the judge decided to send this case to a jury.

OpenAI claims that these diaries have been "taken out of context" by Musk's legal team.

But anyway, next week in court, Brockman will have to face every word he wrote.

A court session that will rewrite the narrative of the AI industry

Looking back at these four days of court sessions, the amount of information is astonishing.

Musk admitted to distillation, breaking the biggest open secret in Silicon Valley. He ranked AI companies and downgraded his own xAI to the smallest in court.

Musk admitted that the Grok model distilled ChatGPT

He was taught by the judge that "you're not a lawyer" and was prohibited from discussing the AI doomsday theory.

His financial steward, Birchall, confirmed the details of the $38 million donation but also exposed the fact that the donation might not have had any restrictions.

And the most dramatic reversal is yet to come. Brockman's witness stand will be the real climax of this drama.

Next Monday, the face - off will officially begin.

Reference materials:

https://www.theverge.com/ai - artificial - intelligence/921546/elon - musk - xai - openai - trial - model - distillation

https://www.cnbc.com/2026/04/30/openai - trial - elon - musk - sam - altman - live - updates.html

https://www.businessinsider.com/takeaways - elon - musk - sam - altman - openai - federal - trial - 2026 - 4?utm_campaign=business - link - post&utm_medium=social&utm_source=twitter

This article is from the WeChat official account "New Intelligence Yuan", author: New Intelligence Yuan. Republished by 36Kr with permission.