Y Combinator CEO urges US open-weight labs to adopt distillation, rejects crackdown

In an interview with CNBC earlier in September 2026, Y Combinator CEO Garry Tan said US regulators should not intervene in Chinese AI labs' use of distillation, and argued that American open-weight AI labs should adopt similar techniques to build a stronger non-Chinese open-weight model ecosystem.

The remarks contrast with Anthropic's position. The company released its second report in September 2026 alleging Chinese labs have engaged in “illicit distillation attacks,” and its CEO Dario Amodei has publicly called on US regulators to crack down on the practice.

Distillation occurs when a model maker extensively prompts another model to learn how it works and reasons, and it is commonly used by AI labs to help train new models. Tan clarified that he is not advocating for US labs to use stolen credentials to distill; he wants them to be free to do so through proper access.

Tan also criticized proprietary AI labs for training on large amounts of human knowledge, including copyrighted material, without permission. He called restrictions on how customers use API calls “overreach,” suggesting the government could help normalize the idea that intelligence trained on broad public data should be more of a public good than locked behind restrictive terms of service.

He wants a balance between open-weight and frontier labs, saying frontier labs should remain fundable and viable as a business model while open-weight models give people freedom and access. In his view, the true “doomer scenario” for AI would be one powerful monolithic company controlling all frontier capabilities.