Musk Testifies xAI Distilled OpenAI Models for Grok

Elon Musk testified Thursday in a California federal court that his artificial intelligence company, xAI, partly used OpenAI models through a technique known as distillation to train its Grok chatbot. The admission came during cross-examination in Musk's ongoing lawsuit against OpenAI, its CEO Sam Altman, and co-founder Greg Brockman, which alleges the company abandoned its original nonprofit mission by shifting to a for-profit structure.

The trial, which began this week, centers on OpenAI's governance and its 2015 founding as a nonprofit aimed at developing AI for humanity's benefit. Musk co-founded the organization that year alongside Altman, Brockman, and others but departed in 2018. When OpenAI lawyer William Savitt asked if xAI had used distillation on OpenAI models, Musk replied, "Partly," adding that "generally all the AI companies do that" and describing it as "standard practice to use other AIs to validate your AI."

Distillation involves querying a larger AI model's public interface or API to generate responses, which are then used as training data for a smaller, more efficient model. This method allows developers to replicate capabilities at lower cost without massive...