Musk Says xAI Used OpenAI Models to Train Grok

In a surprising courtroom moment, Elon Musk confirmed that his AI company, xAI, has partially relied on techniques involving OpenAI models to train its chatbot Grok.

The revelation came during testimony in a California federal court, where Musk is currently suing OpenAI and its leadership, including CEO Sam Altman and co-founder Greg Brockman. Musk alleges the company strayed from its original nonprofit mission by transitioning into a for-profit structure.

Distillation: A Common Industry Practice?

At the center of the discussion is a method known as “distillation.” This involves training new AI models by querying existing ones—essentially learning from their outputs.

When asked directly whether xAI used OpenAI’s models in this way, Musk responded, “Partly,” adding that such practices are common across the AI industry.

While not necessarily illegal, distillation can violate the terms of service set by AI providers. It also raises concerns about fairness, especially given the massive investments companies make in computing power and data.

Rising Tensions in the AI Race

The issue of distillation has become a hot topic among major AI players. Companies like Anthropic and OpenAI have recently taken a stronger stance against third parties using their models to build competing systems.

Much of the focus has been on Chinese firms, which are reportedly using distillation to create powerful, low-cost alternatives to U.S.-based AI models. However, Musk’s testimony suggests that similar practices may also be happening closer to home.

A Bit of Irony

Musk’s admission adds an ironic twist to the broader AI debate. Many leading AI labs have themselves faced criticism for how they collect training data, including potential copyright concerns. Now, those same companies are working to prevent others from learning from their systems.

To address the issue, OpenAI, Anthropic, and Google have reportedly joined forces through the Frontier Model Forum to detect and block suspicious usage patterns, such as large-scale automated queries.

Where xAI Stands

Later in his testimony, Musk offered his view of the current AI landscape. He ranked Anthropic as the leading AI company, followed by OpenAI and Google. He described xAI as a smaller player, noting it has only a few hundred employees compared to its larger rivals.

What It Means Going Forward

Musk’s comments highlight just how competitive—and complex—the AI race has become. As companies push to build smarter models faster, questions around data use, fairness, and intellectual property are becoming harder to ignore.

With legal battles underway and industry tensions rising, the rules of AI development may soon face closer scrutiny.

Also read : Google TV Gets Gemini AI Tools and YouTube Shorts Feed