On Thursday, during a high-stakes federal court testimony, Elon Musk made intriguing remarks that may have significant implications for the ongoing legal tussle between his AI venture, xAI, and the tech behemoth OpenAI. As Musk sat on the witness stand, the lines between collaboration and competition in the fast-evolving AI landscape became even more blurred.
During cross-examination by OpenAI attorney William Savitt, Musk discussed the process known as “distillation.” Savitt asked Musk directly if he was familiar with the term, to which Musk succinctly defined it: “It means to use one AI model to train another AI model.” Musk’s response hints at a broader industry practice, suggesting that xAI could have utilized OpenAI’s models in its own training processes.
When pressed further on the relationship between xAI and OpenAI’s technology, Musk replied, “Generally all the AI companies [do that].” This indicates a prevailing industry norm, where AI developers validate and enhance their models using technologies developed by competitors. Musk’s elucidation on distillation points to a common strategy among AI companies: enhancing efficiency and performance by leveraging existing models’ capabilities.
Distillation, which Musk noted, is more than just an academic concept; it’s a practical technique where a smaller, more efficient AI model is trained to replicate the performance of a larger, advanced model. This not only saves on computational resources but also accelerates deployment, making it a key strategy in the competitive AI arena.
However, the dialogue didn’t stop there. Savitt pressed Musk on whether xAI had used OpenAI’s technology in any capacity to develop its own models. Musk asserted that it is standard practice for AI entities to utilize peer technologies for validation purposes. His answer—“It is standard practice to use other AIs to validate your AI”—suggests an acknowledgment of collaborative aspects often overshadowed by the competitive nature of the AI industry.
This courtroom exchange reflects broader tensions within the AI sector. OpenAI has been actively trying to protect its models from unauthorized distillation, particularly in light of concerns over foreign entities, notably the Chinese lab DeepSeek, replicating American innovations. In a February 2026 memo, OpenAI underscored its commitment to safeguarding its models, reinforcing its competitive stance in a landscape riddled with potential threats from abroad.
The U.S. government has echoed similar concerns. In April 2026, Michael Kratsios from the White House’s office of science and technology policy announced initiatives to inform American AI companies about potential foreign distillation tactics, reaffirming the government’s commitment to a secure and competitive AI environment.
Within this competitive backdrop, American AI labs have also navigated complex relationships. While collaboration can benefit technological advancement, rivalries have led to strict boundaries. For instance, in August 2025, Anthropic blocked OpenAI’s access to its Claude coding models due to alleged violations of terms of service, and more recently, Anthropic has also restricted xAI’s access to its own coding models. These actions highlight a shifting dynamic where competitive interests often take precedence over collaborative efforts.
During the cross-examination, Savitt scrutinized Musk’s historical aspirations regarding OpenAI, questioning his alleged attempts to manipulate funding and researcher recruitment to exert control over the organization. Musk’s history with OpenAI has been highly contentious, and these lines of questioning reveal deep-seated rivalries in the AI space, as each company vies for dominance.
As the legal proceedings continue to unfold, the implications of Musk’s testimony resonate with industry watchers, emphasizing the intricate dance between collaboration and competition that defines the AI landscape today. With tension mounting and interests at stake, the outcomes of this case could have far-reaching consequences for both Musk’s xAI and OpenAI, highlighting the complexities of innovation, protection, and advancements in artificial intelligence.
Inspired by: Source
