While testifying on Thursday in federal courtroom, Elon Musk appeared to point that his AI lab could have used OpenAI’s fashions to coach xAI’s personal. He touched upon the subject whereas sitting on the witness stand answering cross-examination questions from an OpenAI legal professional amid his ongoing authorized battle towards the ChatGPT-maker.
This is the alternate, as greatest as WIRED may seize it:
OpenAI Lawyer William Savitt: Do you already know what distillation is?
Musk: It means to make use of one AI mannequin to coach one other AI mannequin.
Savit: Has xAI accomplished that with OpenAI?
Musk: Generally all of the AI corporations [do that].
Savit: So that’s a sure.
Musk: Partly.
Distillation is a method the place a smaller AI mannequin is educated to imitate the habits of a bigger, extra succesful mannequin, making it cheaper and sooner to run whereas preserving a lot of its efficiency.
OpenAI’s lawyer, William Savitt, then requested whether or not OpenAI’s know-how had been utilized in any approach to develop xAI.
Savit: Has OpenAI know-how been utilized in any approach to develop xAI?
Musk: It is normal observe to make use of different AIs to validate your AI.
OpenAI and xAI didn’t instantly reply to WIRED’s request for remark.
OpenAI has been making an attempt to stop its rivals from distilling its AI fashions, particularly, the Chinese AI lab DeepSeek. In a February 2026 memo to a House committee, OpenAI wrote that it has “taken steps to protect and harden our models against distillation.” In that memo, OpenAI mentioned it was targeted on guaranteeing a taking part in subject wherein “China can’t advance autocratic AI by appropriating and repackaging American innovation.”
The Trump administration has additionally taken steps to stop Chinese corporations from distilling American AI fashions. Michael Kratsios, the White House’s director of the workplace of science and know-how coverage, mentioned in an April 2026 memo that it could share info with US AI corporations about overseas distillation. Kratsios mentioned in a put up on X that the “US government is committed to the free and fair development of AI technologies across a competitive ecosystem.”
American AI labs have used one another’s AI fashions in different methods, to check progress and assess security. But in at present’s aggressive panorama, some AI corporations have fully reduce off rival labs. In August 2025, Anthropic blocked OpenAI’s entry to its Claude coding fashions after the corporate alleged that its phrases of service had been violated. More lately, Anthropic reduce off xAI from utilizing its AI fashions for coding as effectively.
In his multiday cross-examination of Musk, Savitt has questioned Musk about his makes an attempt to imagine management of OpenAI and, subsequently, his quest to beat the ChatGPT-maker. On Wednesday, Savitt offered emails and texts from 2017 to assist a line of questions as as to if Musk squeezed OpenAI by withholding funding and hiring away key researchers.
https://www.wired.com/story/elon-musk-distill-openai-models-partly-xai/