Musk admits xAI used OpenAI models to build Grok
What's the story
Elon Musk, a key player in the artificial intelligence (AI) industry, has admitted that his company may have used models from OpenAI to train its own. The revelation came during Musk's testimony in a federal court on Thursday as part of his ongoing legal battle against OpenAI and its CEO Sam Altman.
AI training technique
Musk says all companies use AI distillation
During his testimony, Musk was asked by OpenAI's attorney if he knew what distillation was. "It means to use one AI model to train another AI model," he said. When asked if xAI had used OpenAI's models for training, Musk said, "Generally all AI companies [do that]." The process of distillation involves training a smaller, less powerful model to replicate the behavior of a larger, more powerful one. This makes it cheaper and faster while retaining much of its performance.
AI model protection
OpenAI's efforts to protect models
OpenAI has been working to prevent other companies from distilling its AI models, especially Chinese lab DeepSeek. In a February 2026 memo to a House committee, OpenAI said it has "taken steps to protect and harden our models against distillation." The company is focused on maintaining a level playing field where "China can't advance autocratic AI by appropriating and repackaging American innovation."
Government action
US government actions on distillation issue
The Trump administration has also taken steps to stop Chinese companies from distilling American AI models. Michael Kratsios, the White House's director of the office of science and technology policy, said in an April 2026 memo that it would share information with US AI companies about foreign distillation. He emphasized that the "US government is committed to the free and fair development of AI technologies across a competitive ecosystem."