A PYMNTS Company

US Lawmakers Seek to Block Use of Chinese AI Models in Federal Agencies

By  |  June 25, 2025

A bipartisan coalition in Congress introduced legislation Wednesday aimed at prohibiting the use of artificial intelligence systems developed in adversarial nations, including China, across U.S. executive agencies. The proposal reflects growing concern among U.S. officials about the national security implications of foreign-developed AI technologies.

    Get the Full Story

    Complete the form to unlock this article and enjoy unlimited free access to all PYMNTS content — no additional logins required.

    yesSubscribe to our daily newsletter, PYMNTS Today.

    By completing this form, you agree to receive marketing communications from PYMNTS and to the sharing of your information with our sponsor, if applicable, in accordance with our Privacy Policy and Terms and Conditions.

    The proposed legislation, titled the No Adversarial AI Act, would establish a permanent framework to bar federal agencies from procuring or operating AI models originating in countries such as China, Russia, Iran, and North Korea. The initiative is being led in the House of Representatives by Republican John Moolenaar of Michigan and Democrat Raja Krishnamoorthi of Illinois, who chair and rank, respectively, on the Select Committee on the Chinese Communist Party.

    According to Reuters, the legislative effort comes on the heels of findings by a senior U.S. official, who determined that Chinese AI developer DeepSeek may be supporting Beijing’s military and intelligence apparatus. The same report indicated that DeepSeek has had access to substantial quantities of advanced Nvidia semiconductors, which are critical to training high-performance AI models.

    DeepSeek made headlines in January with the announcement of a large language model said to rival U.S.-based systems like those from OpenAI, but at a fraction of the cost. Following these claims, several American companies and government entities moved to restrict the use of DeepSeek’s tools, citing cybersecurity threats, per Reuters.

    Read more: Will Trump Tip the AI Race in China’s Favor?

    The proposed bill directs the Federal Acquisition Security Council to compile and maintain a dynamic list of AI models originating in adversarial states. Executive branch agencies would be prohibited from acquiring or utilizing any model on that list unless they receive a specific exemption granted by Congress or the Office of Management and Budget.

    Additionally, the legislation includes a mechanism allowing technologies to be removed from the blacklist, provided there is credible evidence that the products are not controlled or influenced by foreign adversaries.

    “The U.S. must draw a hard line: hostile AI systems have no business operating inside our government,” Moolenaar said in a statement. “This legislation creates a permanent firewall to keep adversary AI out of our most sensitive networks—where the cost of compromise is simply too high.”

    In addition to Moolenaar and Krishnamoorthi, the House bill is co-sponsored by Representatives Ritchie Torres, a Democrat from New York, and Darin LaHood, a Republican from Illinois. The Senate version will be spearheaded by Senators Rick Scott (R-FL) and Gary Peters (D-MI).

    Source: Reuters