The U.S. government is reportedly considering measures that would effectively block domestic corporations from using Chinese artificial intelligence (AI) models. As the technological competitiveness of Chinese AI rises quickly, there is a move to tighten regulations on national security grounds.
Axios, a U.S. online media outlet, reported on the 20th, citing multiple sources, that the U.S. government is considering measures to restrict domestic corporations' access to Chinese AI models.
Recently, Chinese AI startup Moonshot AI released the low-cost, high-performance open-source model "Kimi K3," rapidly narrowing the technology and price gap with the United States.
According to government sources, the United States is reviewing a "gradual but sustained" regulatory approach that would pressure corporations using Chinese AI models by applying government procurement rules or placing them on a restricted transactions list.
Another source said there is a move to discourage corporations from using them by highlighting that Chinese AI models carry potential security vulnerabilities and that their use could create problems for control of a corporation and information security.
If included on the restricted transactions list, U.S. corporations would need government approval to use Chinese AI models, effectively blocking access.
The U.S. government had initially considered an executive order holding corporations liable for security incidents arising during the use of Chinese AI models, or regulating Chinese AI under the pretext of protecting the domestic supply chain.
At the time, however, the effort stalled amid opposition that excessive regulation could hinder innovation. But Axios said that as the hard-line stance on China has strengthened under the Donald Trump administration and the pace of China's AI advances has quickened, the push to restrict access to Chinese AI is regaining momentum.
Some say the measures aim to keep China in check while fostering the open-source AI ecosystem in the United States, but there are also concerns that, in effect, they could further strengthen the market dominance of U.S. AI corporations such as OpenAI and Anthropic, which have called for regulation of Chinese open-source AI.
David Sacks, co-chairperson of the White House President's Council of Advisors on Science and Technology, said the previous day, "Two closed AI labs that have built an oligopoly want the government to eliminate their open competitors," adding, "It's time for the overwhelming majority of Silicon Valley corporations that recognize the value of open competition to take action."
Chairperson Sacks has repeatedly warned about the risks of so-called "regulatory capture," in which specific corporations or industries exert influence over legislative and regulatory bodies to craft rules that favor them.