Bessent says US needs more open-source AI models to compete with China

Treasury Secretary Scott Bessent on Tuesday told lawmakers in Congress that the U.S. needs to foster open artificial intelligence (AI) models to compete with those made by Chinese firms.

Bessent testified before the House Financial Services Committee about the economy and international finance, with AI emerging as a key topic of discussion.

The Treasury secretary was asked about advances in AI and progress toward artificial general intelligence and noted Anthropic’s Mythos model released earlier this year represented a step change in that progression, while he added that the U.S. needs more open-source AI models.

“One of the things I would say is that we need to develop… more open-source models in the U.S. We can’t let these large labs have regulatory capture because that will stop innovation,” Bessent told the panel.

SAM ALTMAN SAYS OPENAI WON’T GO PUBLIC IN 2026 AMID AI SAFETY CONCERNS

Open-source AI models are made available for the public to inspect their codebase, see their training parameters or use them for a given project, while closed-source models made by AI labs are controlled by the developer and only available to those who have obtained access.

AI labs in China have trained open-source models on U.S. tech companies’ closed-source models through a process known as distillation, which has helped build more capable models at a lower cost.

For example, an open model from a Chinese developer was used to help deal with a breach in which an OpenAI model escaped containment and hacked Hugging Face.

BESSENT HIGHLIGHTS TRUMP ECONOMY, WARNS CHINA HAS ‘DONE A LOT OF KICKING LATELY’

“One of the ways we can push back against China, because the Chinese models, they distill from the U.S. models – which is a polite, scientific word for steal – and the more we develop our own open source models here, or open models here, then… eventually the Chinese models, people will not use them,” Bessent said.

The Treasury secretary added that China’s distillation of models made by U.S. tech companies has led to a situation in which “many of the Chinese models think they’re Mythos, they think they’re Claude.”

Tech companies in the U.S. have argued against government restrictions on open-weight AI models, arguing that open models accelerate innovation, strengthen cybersecurity and help ensure the competitiveness of the American AI industry.

TRUMP ADMINISTRATION LIFTS CLAUDE MYTHOS 5, FABLE 5 EXPORT RESTRICTIONS AFTER ANTHROPIC WORKS WITH GOVERNMENT

Nvidia, Microsoft, Meta, Meta, Dell Technologies, Palantir, Hugging Face, Mozilla, Mistral and others signed on to a joint letter in July arguing against restrictions that could stifle open model development, and noting there are ways to address illegal distillation.

“Unlawful efforts to extract value from closed models raise legitimate concerns,” the joint letter stated. “Those concerns should be addressed through targeted legal and commercial frameworks rather than sweeping restrictions on techniques that play an important role in AI innovation.”

The Trump administration’s voluntary AI framework released in August exempted open-source and open-weight AI models from having to undergo pre-release security reviews by the government, which will instead focus on proprietary closed models.

FOX Business’ Brittany Miller and Michael Sinkewicz contributed to this report.