News - Musk Hints at xAI's Use of OpenAI Models in Federal Court Testimony

In a heated federal court session, tech magnate Elon Musk seemingly acknowledged that xAI, his artificial intelligence initiative, had utilized OpenAI's models to enhance its own technology. This revelation came during a rigorous cross-examination led by OpenAI attorney, William Savitt, amidst an ongoing legal conflict between Musk and the creators of ChatGPT. During the exchange, Musk was questioned about the controversial distillation process, where a smaller, more efficient AI model is crafted to mirror the performance of a larger, superior one. Savitt pressed Musk on whether this practice was applied at xAI using OpenAI's models, to which Musk conceded that it is a widespread industry practice, albeit partially confirming xAI's participation. The implications of this conversation extend beyond the courtroom. OpenAI has been meticulously guarding its models against unauthorized distillation by rival labs and has raised these concerns especially about the Chinese AI entity, DeepSeek. In a memo circulated to a House committee in February 2026, OpenAI highlighted its efforts to safeguard its innovations from being repurposed, with an emphasis on countering advancements by autocratic AI through foreign appropriation. Simultaneously, U.S. government officials, including Michael Kratsios, the head of the Office of Science and Technology Policy, have taken stringent measures to restrict the dissemination of American AI knowledge to foreign bodies. Kratsios's communication underscores a commitment to fostering fair competition and protecting intellectual property within the AI sector. The rivalry within American AI labs is intensifying, leading to instances where companies have severed ties to protect proprietary technology. Notably, Anthropic disrupted OpenAI's access to Claude coding models alleging violations of their service terms, and subsequently, severed xAI's access as well. Savitt's examination of Musk over the course of several days revealed elements of Musk's strategic moves to potentially dominate and outmaneuver OpenAI. Evidence presented, including emails and texts from 2017, underpinned the narrative of Musk's calculated endeavor to undermine OpenAI by constricting financial support and pilfering pivotal researchers.