Gensler Flags Chinese AI Models as Risk to U.S. Data-Center Returns
The former SEC chair said lower-cost, open-weight models could weaken U.S. pricing power as companies spend heavily on AI infrastructure.
Gary Gensler warned that lower-cost Chinese artificial intelligence models could spread faster than U.S. systems, putting pressure on American companies’ pricing power and data-center returns.
Gensler, former chair of the U.S. Securities and Exchange Commission and an MIT professor, said Chinese open-weight models may diffuse through China’s economy and become standardized more quickly than the largely closed large language models offered by U.S. companies.
“It’s financially not sustainable. You cannot spend $750 billion a year building data centers and only bring in $100 to 150 billion in revenues ...” Gensler said.
Lower-cost models that perform well enough for common tasks could reduce the pricing power of U.S. developers and make it harder for companies to earn adequate returns on heavy infrastructure spending. Gensler’s comments frame the competition with China as both a technology contest and a financial question for the U.S. AI industry.
“Unlike Chinese models, which are open weight which may diffuse AI technology through their economy and standardize more quickly, American model companies are largely offering closed large language models,” Gensler said.
Chinese models were estimated to account for 1% of global AI workloads in late 2024 and 30% by the end of 2025. A separate 2025 measure of token traffic on one platform found Chinese open-weight models rising from a weekly low of 1.2% in late 2024 to nearly 30% of usage in some weeks. The two measures cover different activity and are not directly comparable.
The House Committee on Homeland Security and the House Select Committee on China announced a joint investigation April 29, 2026, into national-security and cybersecurity risks involving models developed by DeepSeek, Alibaba, Moonshot AI and MiniMax.
The committees raised allegations involving possible unauthorized distillation of American AI systems through proxy accounts, efforts to evade access restrictions and potential violations of companies’ terms of service. The committees also said model distillation can be legitimate and did not establish that every named company engaged in the alleged conduct.
Gensler said AI is already changing financial activity, including automobile underwriting, insurance underwriting and high-frequency trading. “It’s not just how AI is influencing finance—and it is—it’s changing how we do automobile underwriting, insurance underwriting, and how we trade: high frequency trading more than how we individually trade,” he said.
