In a significant move to safeguard national security, a bipartisan coalition of U.S. lawmakers has introduced the No Adversarial AI Act, aiming to ban the use of artificial intelligence (AI) systems developed by nations deemed adversarial, including China, Russia, Iran, and North Korea, within federal agencies. This legislative initiative underscores growing concerns over the potential risks associated with foreign-developed AI technologies infiltrating critical government operations.
Background and Legislative Intent
The rapid advancement of AI technologies has intensified global competition, particularly between the United States and China. Recent developments, such as the emergence of Chinese AI firm DeepSeek, have heightened apprehensions. DeepSeek has reportedly developed AI models that rival leading U.S. counterparts at a fraction of the cost, raising alarms about China’s accelerating capabilities in this strategic domain.
Representative John Moolenaar, a Republican from Michigan and chairman of the House Select Committee on China, emphasized the gravity of the situation, stating, We are in a new Cold War, and AI is the strategic technology at the center. He further highlighted concerns over the Chinese Communist Party’s (CCP) approach to AI, which he characterized as focusing on theft, scaling, and subversion rather than innovation. Moolenaar stressed the necessity of drawing a clear line to prevent U.S. government systems from being powered by tools designed to serve authoritarian interests.
Key Provisions of the No Adversarial AI Act
The proposed legislation outlines several critical measures to mitigate the risks posed by adversarial AI systems:
1. Identification and Listing of Adversarial AI Systems: The Federal Acquisition Security Council (FASC) would be tasked with creating and maintaining a publicly accessible list of AI models developed by foreign adversaries. This list would be updated biannually to ensure it reflects the current landscape of potential threats.
2. Prohibition of Use in Federal Agencies: Executive agencies would be prohibited from acquiring or utilizing AI technologies developed by entities on the FASC’s list. This blanket ban aims to prevent any integration of potentially compromised AI systems into federal operations.
3. Exemptions and Oversight: The legislation allows for narrow exceptions, such as research, testing, or national security purposes. However, any exemption would require explicit approval from Congress or the Office of Management and Budget (OMB), ensuring stringent oversight and accountability.
4. Delisting Process: Companies that can demonstrably prove they are free from foreign adversary control or influence may petition for removal from the FASC’s list. This provision ensures that entities unjustly categorized have a pathway to rectify their status.
Bipartisan Support and National Security Implications
The bill has garnered support from both sides of the political aisle, reflecting a unified stance on the importance of protecting national security interests. Co-sponsors include Representatives Raja Krishnamoorthi (D-IL), Ritchie Torres (D-NY), and Darin LaHood (R-IL), as well as Senators Rick Scott (R-FL) and Gary Peters (D-MI).
Representative Krishnamoorthi underscored the risks associated with adversary-controlled AI, stating, Artificial intelligence controlled by foreign adversaries poses a direct threat to our national security, our data, and our government operations. He emphasized that the bipartisan legislation aims to establish a clear firewall between foreign adversary AI and the U.S. government, thereby protecting American institutions and citizens.
Contextualizing the AI Competition
The introduction of this legislation comes amid a broader context of intensifying technological rivalry between the U.S. and China. The 2025 AI Index Report from Stanford University’s Human-Centered Artificial Intelligence center indicates that while the U.S. currently leads in producing top AI models, China is rapidly closing the performance gap. The report notes that China reached near parity with the U.S. in 2024 on several major benchmarks and leads in AI publications and patents.
Experts have highlighted the strategic importance of AI in this new era of competition. Jack Clark, co-founder and head of policy at Anthropic, pointed out that AI systems inherently reflect the values of the societies that develop them. He warned that AI built in authoritarian nations would be inescapably intertwined and imbued with authoritarianism, underscoring the need for decisive action to ensure that democratic values prevail in the development and deployment of AI technologies.
Broader Legislative Efforts and Export Controls
The No Adversarial AI Act is part of a series of legislative efforts aimed at addressing the multifaceted challenges posed by foreign adversaries in the realm of advanced technologies. Earlier this year, lawmakers introduced the Chip Security Act to tackle the smuggling of export-controlled AI chips, particularly to China. This bill mandates AI chip manufacturers to incorporate technology to verify chip locations before exports, aiming to prevent advanced AI technology from being accessed by potentially hostile entities.
Additionally, there have been proposals to task the National Security Agency (NSA) with developing an AI security playbook to counter threats from nations like China. This initiative would focus on identifying vulnerabilities in advanced AI data centers and among developers capable of producing critical AI technologies, with an emphasis on cybersecurity risks unique to protecting such technologies.
State-Level Actions and International Context
Beyond federal initiatives, several U.S. states have taken proactive measures to mitigate risks associated with foreign-developed AI applications. For instance, New York, Virginia, and Texas have banned the use of the Chinese AI app DeepSeek on state-owned devices and networks. These actions reflect a growing recognition at multiple levels of government of the potential threats posed by adversarial AI systems.
Internationally, other countries have also begun to address similar concerns. Nations such as Italy, Taiwan, South Korea, and Australia have implemented or are considering restrictions on the use of certain foreign-developed AI applications within their governmental operations, highlighting a global awareness of the strategic implications of AI technologies.
Conclusion
The introduction of the No Adversarial AI Act represents a concerted effort by U.S. lawmakers to fortify national security by preventing the integration of potentially compromised AI systems into federal agencies. By establishing a framework to identify, list, and prohibit the use of AI technologies developed by foreign adversaries, the legislation aims to protect sensitive government operations from external influence and potential espionage.
As the global competition in AI continues to intensify, such legislative measures underscore the importance of maintaining technological leadership and safeguarding democratic values in the development and deployment of artificial intelligence. The bipartisan nature of this initiative reflects a shared commitment to addressing the complex challenges posed by the rapid advancement of AI technologies on the international stage.
