The DeepSeek-R1 artificial intelligence model has been a wake-up call for the U.S. technology sector, prompting calls for stricter export controls to curb China’s access to advanced technologies. But as AI increasingly redefines global power—through its proven use in battlefield decision-making, mass surveillance, automated cyberoperations, and application of state propaganda—Washington’s strategy must also increase investment in research and development, foster domestic innovation, and enhance international collaboration to promote AI that is free from state-imposed censorship and aligned with democratic values.
In the race for AI supremacy, the finish line isn’t just economic dominance—it’s the future of freedom. The United States must not only maintain its technological edge but also promote an international AI ecosystem rooted in transparency, openness, and democratic governance. AI can either reinforce repression or empower free societies. It must be developed and deployed in ways that preserve civil liberties, protect open debate, and resist ideological conformity.
Unlike U.S. AI models that prioritize open knowledge and free expression, albeit imperfectly and under increasing scrutiny, China’s AI platforms serve as instruments of authoritarian control. While U.S. models face legitimate criticism over bias, data sourcing, and political pressure in academia, they are still developed in an open environment subject to legal challenge, public transparency, and independent oversight—conditions that do not exist under China’s centralized censorship regime. The United States must not only maintain its technological edge but also actively promote U.S. AI as the superior alternative. Beijing isn’t building smarter machines—it’s building better censors.
Initiatives such as the AI Infrastructure Partnership—a collaboration among xAI, Nvidia, Microsoft, BlackRock, and Abu Dhabi AI investment fund MGX—point the way. While not all members operate within democratic systems, the infrastructure, governance, development, and deployment of AI through these partnerships are still governed by the U.S. rule of law, transparency standards, and regulatory accountability. These projects are subject to congressional oversight, disclosure requirements from the U.S. Securities and Exchange Commission, and scrutiny from independent media and civil society watchdogs—mechanisms largely absent in authoritarian systems. Trust in these initiatives should rest not on corporate benevolence but on enforceable norms, legal safeguards, and multilateral cooperation that distinguish them from authoritarian models. The United States must lead not only with innovation but with an invitation to collaborate with like-minded partners.
At the AI Action Summit in Paris, U.S. Vice President J.D. Vance reaffirmed Washington’s commitment to technological dominance, stating that the United States is “the leader in AI and our administration plans to keep it that way.” He emphasized that U.S. AI leadership is built on advanced semiconductor design, frontier algorithms, and transformational applications.
Vance also warned of the dangers posed by hostile foreign adversaries that weaponize AI for mass surveillance, censorship, and historical revisionism. His remarks were widely seen as a rebuke of China’s AI-driven censorship, exemplified by DeepSeek’s suppression of discussions on topics such as the Tiananmen Square massacre and the genocide and forced labor of Uyghurs.
Former Google CEO Eric Schmidt underscores the urgency for the United States to adapt its AI strategies in response to China’s rapid advancements. He warns in a recent policy paper with Alexandr Wang and Dan Hendrycks that China’s technological gap with the United States is narrowing quickly and that Beijing’s aggressive investment in open-source AI could erode U.S. advantages. Schmidt cautions against pursuing an AI “Manhattan Project” approach that could escalate tensions, instead advocating for robust energy infrastructure and smarter policy to support open, competitive development. These insights reinforce the strategic importance of not just technological acceleration but governance, openness, and global trust in U.S. AI.
This strategic commitment is reflected in the Trump administration’s America First trade policy, which mandates a comprehensive review of U.S. export control systems. The goal is to close loopholes that enable adversaries to access critical technologies while reinforcing enforcement mechanisms to prevent unauthorized technology transfers. For example, in 2023, the U.S. Commerce Department expanded the Entity List to include additional Chinese firms tied to military and surveillance programs. Additionally, export restrictions were imposed on Nvidia’s A100 and H100 chips to prevent their shipment to China, underscoring how targeted controls can deny strategic computing resources without blanket bans.
Trump has further underscored the strategic importance of AI through actions such as an executive order on U.S. leadership in AI, which states: “It is the policy of the United States to sustain and enhance America’s global AI dominance in order to promote human flourishing, economic competitiveness, and national security.”
U.S. leadership in AI and semiconductors depends on a carefully calibrated export control strategy that balances national security with market competitiveness. Rather than solely claiming that U.S. AI promotes democracy, the more durable policy goals should be to preserve and expand U.S. leadership in AI ecosystems that operate under transparent governance, regulatory oversight, and international standards that are inconsistent with authoritarian control.
In January, the Biden administration introduced the Framework for Artificial Intelligence Diffusion interim final rule (IFR) to block Chinese access to advanced AI chips. The rule set up a tiered system that limits AI chip exports to specific countries while capping shipments to others. However, critics—including the Semiconductor Industry Association, compliance experts, and national security analysts—argue that this system is unrealistic and unenforceable, restricting U.S. companies from selling chips in key markets. Export controls shouldn’t constrain allies; they should constrain adversaries.
Even NATO allies such as Latvia, Estonia, and Lithuania expressed concerns, arguing that the rule “risks creating artificial divisions in the EU Single Market and undermine[s] AI ecosystem development.” They called on the United States to reconsider the IFR’s impact on allied innovation, warning that it could jeopardize trust and collaboration within trans-Atlantic tech ecosystems. By restricting access to U.S. chips, the IFR inadvertently incentivizes foreign nations to invest in non-U.S. technology alternatives, weakening America’s long-term economic influence.
The rule has faced not only foreign criticism but also domestic pushback. Seven Republican senators and several U.S. tech firms—including Nvidia and Oracle—have warned that the tiered framework is overly complex and risks driving allied countries to adopt Chinese AI alternatives. The Trump administration is reportedly exploring a replacement for the current tier system, possibly through a streamlined, government-to-government licensing model that aligns with broader trade strategy. Another proposed revision would lower the chip shipment exception threshold from 1,700 to 500 units of Nvidia’s H100—an effort to tighten enforcement without imposing blanket restrictions. These discussions highlight a growing bipartisan consensus: that while export controls are necessary, they must be designed to strengthen U.S. competitiveness and avoid alienating trusted partners.
One serious unintended consequence of the IFR is that nations unable to access superior U.S. AI chips could be forced to rely on inferior Chinese alternatives, such as Huawei’s AI processors. Huawei’s chips are significantly less powerful than Nvidia’s, offering a poor substitute for high-performance computing. This shift would not only reduce global AI development capacity but also increase dependence on Chinese technology, further empowering Beijing’s digital influence. Instead of isolating China, the rule could fragment the global AI ecosystem, making it more difficult for the United States to maintain technological leadership.
Washington must reassess and refine its approach to export controls to ensure that key allies and global markets remain reliant on U.S. AI technology rather than inadvertently pushing them into China’s sphere of influence.
In retaliation, China has accelerated its domestic semiconductor production and imposed export controls on rare-earth minerals such as gallium and germanium—critical materials for semiconductor manufacturing. Despite U.S. efforts, Huawei has continued producing advanced AI chips, demonstrating the difficulty of isolating China from global supply chains.
The ongoing U.S.-China AI chip conflict is expected to escalate further, with Washington considering additional restrictions on AI model training data and cloud computing infrastructure. This intensifying tech rivalry underscores the need for the United States to refine its export control strategy while avoiding policies that inadvertently weaken its own industries.
To maintain its global edge, the United States must close enforcement loopholes to prevent adversarial nations from acquiring sensitive technologies, enhance regulatory oversight while penalizing companies that knowingly facilitate China’s AI-driven repression, expand investment in AI and semiconductor R&D to sustain technological leadership, and strengthen international collaboration with allies to develop a unified export control framework that aligns with shared security interests.
At the same time, the United States must actively promote U.S. AI as the global standard, ensuring that businesses and governments worldwide choose AI models that uphold free expression and democratic values—not those designed to censor and suppress information. This effort must also confront valid concerns from U.S. allies, especially in Europe, that have observed rising pressure on U.S. institutions to conform ideologically. While recent domestic trends—such as political influence over universities and legal professionals—raise red flags, the key distinction lies in the U.S. system’s ability to self-correct. Judicial review, a free press, and independent research institutions provide guardrails that authoritarian regimes intentionally suppress. To retain global trust and credibility, the United States must lead by example, governing AI with transparency, accountability, and a commitment to open discourse.
This isn’t just about chips; it’s about choosing sides in the fight for digital freedom. The long-term goal should be global AI adoption that aligns with open and transparent governance, rather than authoritarian control.
AI is not just a tool of economic progress but a geopolitical weapon. The United States must act decisively to lead AI innovation, secure its technological edge, and ensure that AI serves democratic values rather than authoritarian control. If Washington fails to act strategically, it risks ceding its technological and economic primacy to Beijing. Every delay is a gift to China.
The time to act is now.
The post U.S. AI Leadership Needs Smarter Controls appeared first on Foreign Policy.
