Missouri Sen. Josh Hawley introduced legislation on Feb. 4, 2025, that may change the future of AI development globally. Hawley’s bill would prohibit Americans’ use of, investment in or participation in the development of Chinese AI. The bill, titled Decoupling America’s AI Capabilities from China Act of 2025 (DAAICCA), would prohibit American citizens and companies from advancing AI capabilities within the Republic of China and for other purposes.
DAAICCA was introduced days after Chinese developer DeepSeek released an AI language model, the “r1.” This model has garnered attention in the technology sector due to its competitive algorithm, which has been reported to match existing AI technology on the market, the purported low development cost and, critically, its open-source nature that allows anyone to freely train and create an AI model using the r1 language model. Baidu, a leading Chinese internet search engine, has since released an AI-reasoning model to compete with the r1 model.
Hawley, a member of the Senate Judiciary Subcommittee on Privacy, Technology, and the Law, has previously highlighted international AI issues during congressional hearings. He emphasized that while maintaining competitiveness with China in AI development is important, protecting the American public from potential AI-related harms is paramount. In a statement released alongside the proposed bill, Hawley reiterated his concerns about the consequences of Chinese advancements in AI for Americans: “Every dollar and gig of data that flows into Chinese AI are dollars and data that will ultimately be used against the United States. … America cannot afford to empower our greatest adversary at the expense of our own strength. Ensuring American economic superiority means cutting China off from American ingenuity and halting the subsidization of CCP innovation.”
He is not the only Republican who has identified such issues. On Feb. 14, 2025, Texas Attorney General Ken Paxton publicly notified DeepSeek that its platform violates the Texas Data Privacy and Security Act and announced an investigation into the company’s data retention practices. The investigation remains ongoing.
The issue has also attracted bipartisan attention. Democratic Sen. Elizabeth Warren of Massachusetts coauthored a letter with Hawley directed to President Donald Trump’s then-nominee for commerce secretary, Howard Lutnick, urging Lutnick to take swift action in response to the “breakthrough development of an advanced AI model by a Chinese firm.” The Feb. 3 letter outlines numerous national security concerns presented by China and urges application of stronger export control on technological components that, in their words, were necessary to the creation and training of DeepSeek’s AI model.
Critics have expressed concern over what they perceive as DAAICCA’s overly broad definitions, sweeping prohibitions and harsh penalties. For instance, the bill defines AI as “any system capable of performing tasks such as decision-making, learning from experience, and generating human-like outputs.” This definition could arguably encompass any computer programs capable of predicting outcomes, including basic search engines.
The bill primarily targets three AI-based interactions between American citizens and companies and various individuals, groups and entities in China:
- Conducting, aiding in or transferring research and development within China or for, on behalf of or in collaboration with entities of concern. The bill broadly defines “entities of concern” as “any entity, including corporations, universities, and research institutions, organized under the laws of the PRC, has its headquarters or principal place of business in China, or is affiliated with the Chinese Communist Party or the People’s Liberation Army.”
- Exchange of technology or intellectual property related to AI. Another section of DAAICCA prohibits American individuals and entities from importing AI technology or intellectual property developed or created in China. Reciprocally, the bill prohibits Americans from exporting any AI technology or intellectual property to Chinese entities of concern or within China. Notably, the bill does not carve out the exchange of information between American citizens within China. Conceptually, DAAICCA would prevent employees of American companies with locations in China from participating or using AI technology or intellectual property used elsewhere.
- Holding an interest in Chinese entities developing AI. A separate provision prohibits American entities and individuals from “holding or managing any interest in or lending money or extending credit to” Chinese entities involved in the development of AI-related research. “Interest” is defined as direct and indirect ownership in or contractual arrangement with a Chinese entity of concern.
DAAICCA would impose steep penalties. Although the above three provisions are organized under different enforcement mechanisms, violations of any of the provisions can result in fines of up to $100 million for corporations, which would also risk forfeiting any federal “license, contract sub-contract, grant, or public benefit.” Individuals, including officers, directors and employees of violating entities, may face personal fines of up to $1 million and similar loss of federal contracts and grants. Individuals who violate this law would also be ineligible to receive federal financial assistance for five years following the imposition of a penalty, potentially stripping them of the ability to receive federal food assistance programs, emergency housing assistance and student aid.
It remains uncertain whether this bill will pass as written given the criticism. Many critics have pointed out the bill’s potential to stifle AI advancement due to its proposed prohibition on collaboration between already deeply interlinked American and Chinese researchers and developers. Preventing these collaborations has been described as unrealistic for many large companies and organizations that have incorporated AI into their business processes and workflows. Such strict regulation of AI may also contradict the Trump administration’s stated aim of relaxing AI regulation At the Artificial Intelligence Action Summit in France held earlier this year, Vice President JD Vance expressed similar sentiments, stating that he believed excessive regulation in the AI space could “kill a transformative industry just as it’s taking off,” and pledged to refrain from heavy legislating.
For updates on DAAICCA’s progression and coverage of other federal and state regulations on AI, contact the authors of this article.