Testifying in the Musk v. Altman trial in California federal court on April 30, 2026, Elon Musk acknowledged that xAI used distillation techniques on OpenAI models to train Grok, calling the practice "a general practice among AI companies." The admission's legal status is ambiguous—distillation may violate terms of service rather than constitute an explicit legal violation—but it confirms that frontier model capability can be substantially replicated cheaply through distillation, undermining the competitive moat of large compute investments. OpenAI, Anthropic, and Google have begun coordinating through the Frontier Model Forum to detect and limit distillation by monitoring suspicious mass-query patterns.