Y Combinator CEO Garry Tan wants U.S. regulators to stay out of disputes over Chinese AI labs using distillation, and he says American open-weight labs should be free to use the same technique on American frontier models. In a CNBC interview, Tan said ‘I would do nothing’ when asked about regulators stepping in. He elaborated to TechCrunch that there should be ‘an American distillation regime,’ meaning smaller U.S. open-weight AI labs should run the same training techniques on American frontier labs, giving the U.S. a more robust set of open-weight options that aren’t Chinese.

Distillation is when a model maker extensively prompts another model to learn how it works and reasons; it is a commonly and legitimately used training technique. Anthropic this week released a second report claiming Chinese labs are conducting ‘illicit distillation attacks,’ hiding their identities, relying on fraud, and using stolen credentials. Anthropic CEO Dario Amodei had previously called on U.S. regulators to crack down on distillation. Tan’s disagreement is notable given Y Combinator‘s standing as Silicon Valley’s prominent accelerator.

Tan is not defending stolen credentials. He wants open-weight labs to come in through the front door. His argument has two parts: proprietary labs overstep when they decide what customers can do with the information their models share, and the frontier labs did not ask permission when they vacuumed up human knowledge to train models, including plenty of copyrighted material. He told TechCrunch that controlling what users do with API calls to closed-weight models feels constraining, and that access to intelligence trained on broad public data should be more of a public good than something locked away behind restrictive terms of service.

Tan wants a balance between frontier labs and open-weight models. Frontier labs should still be fundable and a great business model, while open-weight models should give people freedom and access. His stated nightmare scenario is a single monolithic AI company that runs away with best access to capital and AI researchers, concentrating all the power of frontier AI in one place.