Washington has made it clear that retaining AI dominance over China is both an economic and national security imperative. In November 2024, the U.S.-China Economic and Security Review Commission recommended a “Manhattan Project-like program” to achieve artificial general intelligence (AGI) before China.¹ Two months later, the Stargate project was announced, mobilizing $500 billion in private capital to strengthen American AI infrastructure and energy production.² President Trump’s January 23 Executive Order left no doubt: “It is the policy of the United States to sustain and enhance America’s global AI dominance.”³

Alongside steps taken to strengthen the domestic AI industry, Washington has also imposed increasingly stringent controls on exports of AI chips and semiconductor manufacturing equipment to China. Initially formulated during President Trump’s first term, these restrictions have since expanded.⁴ The second Trump administration is expected to tighten them further.⁵

These two approaches – strengthening the domestic AI sector and securing key AI resources from Chinese access – form the backbone of America’s AI strategy. Yet, alone, they may not be enough to sustain American leadership in what is arguably the most consequential technology of our time.

Under President Trump, a third pillar must be established: strategic promotion — active efforts to position American AI and compute as the default for international markets critical to U.S. national security and economic competitiveness. This means financing compute infrastructure, providing training and technical assistance for AI adoption at scale, and securing trade agreements that encourage reliance on trusted American AI and cloud providers while enforcing strong security standards.

This approach complements the ‘strengthen and secure’ core of U.S. AI policy by proactively exporting and embedding key American AI technologies in strategically important parts of the world. Done right, strategic promotion can bolster America’s economic power, limit China’s ability to seize new markets, promote democratic values overseas, and forge international partnerships aligned with the U.S. vision for AI leadership and governance.

Why “Strategic Promotion” Matters More Than Ever

Four key imperatives make clear that now is the time to launch a strategic promotion agenda as a complement to strict export controls.

1. Guarding Against China’s “Digital Silk Road”

China’s Belt and Road Initiative has already transformed how infrastructure projects are financed around the world. Its “Digital Silk Road” extends this approach to digital technologies — everything from fiber-optic cables and data centers to cloud services and “smart city” platforms with AI-powered surveillance and data collection capabilities.⁶ By offering steep subsidies, concessional financing, and locally relevant technology and infrastructure solutions, Beijing makes its domestic vendors especially attractive to governments seeking cost-effective routes to modernize their economies.

This playbook has already reshaped global telecommunications. For example, Huawei’s 5G equipment has become ubiquitous across Africa, Asia, and Europe, raising Western security concerns.⁷ In a recent interview, Office of Science and Technology Policy Director Michael Kratsios recalled the “Huawei Wars,” in which the first Trump administration urgently negotiated to remove Huawei telecommunications infrastructure from allied nations due to security concerns.⁸ The lesson from Huawei is clear: waiting until Chinese tech is already embedded in global infrastructure is too late.

Despite U.S. AI superiority, without proactive efforts to promote American AI solutions along the full technology stack — ranging from cloud compute to tailored AI applications that meet local demand — many countries could opt for Chinese offerings purely because they appear cheaper and arrive as comprehensive packages. Even if China lags the United States in chip design, chip manufacturing, and AI development, Beijing can still penetrate strategic global markets by offering just “good enough” AI systems and digital infrastructure at appealing prices and in comprehensive aid packages that extend beyond just compute and AI models.⁹ Many countries in the Global South do not necessarily covet the most advanced AI chips from NVIDIA or access to the absolute frontier of AI capabilities. Rather, more basic commercial AI solutions that can help modernize public services, government functioning, and local economic activity will often suffice. This is especially true if these systems come paired with digital infrastructure aid and technical adoption assistance. These are precisely the kinds of export promotion efforts at which China has proven so effective through its Digital Silk Road initiative.¹⁰

If CCP-backed AI ecosystems take hold, switching to American alternatives will become costly and impractical, especially as Chinese firms improve their AI offerings over time. An offensive approach where the United States steps in first with robust financing, dependable technology, and long-term cloud compute deals, can preempt Chinese efforts to entrench their platforms in key emerging economies.

2. Strengthening America’s Future AI Industry

As AI systems become essential to healthcare, finance, defense, and public services worldwide, early engagement in emerging markets will ensure American technology underpins the next phase of global economic growth. Strategic promotion can expand the commercial base for American AI and compute providers, delivering both immediate economic gains and long-term competitive advantages for American industry.

The immediate gains of establishing a foothold in emerging markets will compound rapidly. Over the next three decades, five African nations, along with Pakistan, the Philippines, and India, will drive more than half of global population growth and see substantial economic expansion.¹¹ These high-growth markets are at an inflection point, choosing how to integrate AI into their economies — a decision that will define their trajectories for decades. Embedding American compute and AI systems as these countries make foundational decisions about their digital and economic futures will not only reinforce U.S. economic strength but also limit Chinese firms’ ability to capture these markets and channel AI-driven profits into their own technological ambitions.

In his speech at the AI Action Summit in Paris, Vice President J.D. Vance pledged the United States would set the global “gold standard” for AI, ensuring that America is “the partner of choice” for countries wanting to expand their own use of AI.¹² Fulfilling this promise requires proactive partnerships to embed American AI in emerging economies.

3. Advancing American Diplomatic and Security Goals

Expanding American AI systems and compute providers into new markets is not just an economic opportunity; it is a geo-strategic necessity. As AI competition with China intensifies, securing market dominance now will determine who sets the global standards for international AI governance and security in the future. In the past, Beijing has at times leveraged subsidized infrastructure assistance to cement favorable diplomatic relationships with recipient nations.¹³ Looking ahead, nations whose digital ecosystems depend on American compute and AI services will have stronger incentives to align with American priorities on key issues such as enforcing export controls, adhering to digital security protocols, and decoupling from Chinese digital infrastructure.

To maximize impact, the United States can pursue government-to-government AI agreements, not just private-sector exports. These agreements can tie AI financing, trade deals, and integration assistance to strong security and governance commitments, such as those described in the AI Diffusion Rule, to limit Beijing’s influence over AI standards worldwide.¹⁴

Coordination with international partners is essential for effective policy implementation. For instance, export controls on advanced AI chips only work if allied nations take steps to ensure U.S. chips are not smuggled out of the country or accessed remotely through the cloud. If the United States denies China access to specific chips, but another partner covertly provides access, then export controls do nothing to hinder China’s technological progress. Likewise, safeguarding key intellectual property, such as model weights for frontier AI systems, depends on shared commitments to secure the data centers where they are housed. When allied economies rely on American AI infrastructure, they have greater incentives to uphold these security commitments.

4. Promoting Democratic Values Via AI Exports

Promoting American AI technology abroad ensures that democratic principles shape the design and deployment of systems central to daily life. As AI becomes more integral to digital life and economic activity, its underlying values will determine how information flows and individual rights are protected. If authoritarian regimes set these standards, we risk censorship, increased surveillance, and an information ecosystem controlled by the CCP.

Concrete examples underscore these concerns. For instance, Chinese developers are required to pass tests that ensure that new models don’t pose security risks, including violating “core socialist values.”¹⁵ When accessed via the cloud, models like those recently developed by DeepSeek restrict their responses to politically sensitive prompts to conform with these values.¹⁶ In addition to censorship concerns, widescale deployment of Chinese models puts user data at risk; by law, a private actor like DeepSeek would need to hand over any data it collected to the Chinese Ministry of State Security, if asked.¹⁷ Recent reporting suggests the CCP is expanding its oversight of DeepSeek, further raising concerns about state control over AI systems and the data they process.¹⁸

By entering growing markets first — executing deals that include financing and on-the-ground support for AI and cloud services adoption — U.S. firms can preempt Chinese offerings and safeguard the democratic principles that guide AI development at home. In doing so, the United States can help establish a global digital landscape where free speech, privacy, and civil liberties are better protected.

Promotion Within an Improved AI Diffusion Framework

Far from rolling back the AI Diffusion Framework in favor of blanket export promotion, the United States could both strengthen its export controls on adversaries and still aggressively promote AI exports to selected partners. Export controls and strategic promotion must work in tandem as part of a comprehensive AI foreign policy strategy. The first denies Beijing direct access to the crown jewels of advanced AI development, while the second secures American leadership in global AI deployment.

The Trump administration can strengthen the AI Diffusion Framework by banning the sale of high-performance inference chips to adversaries (Tier 3 countries) while pushing Congress to increase funding for the Bureau of Industry and Security and crack down on illicit chip smuggling. Simultaneously, numerous options are available for strategic promotion that fit within the bounds of the Diffusion Framework. One example is leveraging the U.S. Export-Import Bank to provide export financing for fundamental digital infrastructure projects, cloud compute access, and tailored AI solutions for emerging markets. In practice, this means offering financing deals to recipient countries to make it easier for them to purchase and implement American technology products along the AI stack.

In 2019, President Trump reauthorized the bank’s mandate, establishing the China and Transformational Exports Program specifically to counter the PRC’s Belt and Road Initiative. However, between 2020-2023, only $8 million was delivered by the bank in financing for AI-related technology exports. This is a negligible amount given China’s investment in AI infrastructure worldwide.¹⁹ Now, the Trump Administration has the opportunity to increase this number significantly. Bureaucratic restrictions limiting the bank’s ability to counter Beijing’s expansion should be lifted, an adjustment the bank’s own president requested in 2024 to improve American competitiveness.²⁰

To compete with China for global deployment leadership, strategic promotion must move beyond just chip or AI model exports and provide comprehensive AI packages. While the CCP pursues AI-powered smart cities development in the Global South, the United States can make even more appealing deals that focus less on surveillance applications and more on targeted local needs. This is where institutions like the U.S. Trade and Development Agency (USTDA) can position American offerings that address partner country priorities, such as healthcare analytics and efficient government services, as the go-to option. Bundling infrastructure and software solutions with implementation assistance and long-term financing for deals with U.S. cloud providers designated as Verified End Users under the Diffusion Framework can help establish friendly AI ecosystems in partner countries while upholding critical security requirements.

No Future Lead Guaranteed

The window for action is open now. Export controls and American innovation have given the United States a lead in every aspect of the AI supply chain, from chip design up to the application layer. This lead is not guaranteed to last forever, even with continued efforts to restrict China’s access to key resources. Many governments around the world are now deciding on the digital infrastructure and AI-related investments that will underpin their economies for decades.

If the United States fails to aggressively promote its AI ecosystems abroad, it risks Beijing’s Digital Silk Road filling the void. But if Washington seizes this moment by strategically exporting and financing AI technologies abroad while maintaining strong controls on adversaries, America can extend its lead, protect its security interests, and shape a world where AI development remains firmly anchored in American innovation.

1. U.S.-China Economic and Security Review Commission, 2024 Annual Report to Congress, 2024. ↩︎
2. “Trump to Announce Private Sector AI Infrastructure Investment – CBS Reports,” Reuters, January 21, 2025. ↩︎
3. White House, Removing Barriers to American Leadership in Artificial Intelligence, January 2025. ↩︎
4. Gregory C. Allen, Press Release: Advanced Computing and Semiconductor Manufacturing Controls Final (Washington DC: U.S. Bureau of Industry and Security, October 7, 2022). ↩︎
5. Emily Feng, “Nvidia Discloses That U.S. Will Limit Sales of Advanced Chips to China after All,” NPR, April 16, 2025. ↩︎
6. Adam Segal, “China’s Digital Silk Road,” Council on Foreign Relations, March 30, 2021. ↩︎
7. Joe Cochrane, “China’s Digital Silk Road in Africa Raises Questions,” Voice of America, March 3, 2023. ↩︎
8. “How to Rebuild the Arsenal of Democracy,” Statecraft, February 5, 2025. ↩︎
9. Colin H. Kahl, “Is America Winning the Race for Global AI Primacy Now?” Foreign Affairs, January 15, 2025. ↩︎
10. “Assessing China’s Digital Silk Road Initiative,” Council on Foreign Relations. ↩︎
11. “The African Century,” Finance & Development, September 2023. ↩︎
12. J.D. Vance, “Speech at Paris AI Action Summit,”  February 11, 2025. ↩︎
13. Noah McBride, et al., “China’s Massive Belt and Road Initiative,” Council on Foreign Relations, January 28, 2020. ↩︎
14. Lennart Heim, “Understanding the Artificial Intelligence Diffusion Framework,” RAND Expert Insights, January 14, 2025. ↩︎
15. Nicholas Welch, “China’s GenAI Content Security Standard: An Explainer.” ChinaTalk, January 7, 2025. ↩︎
16. “Was Zuck Right About Chinese AI Models?” Interconnect, February 1, 2025. ↩︎
17. Bari Weiss, “Can America Win the AI War with China?” The Free Press, February 6, 2025. ↩︎
18. “DeepSeek: China’s National Treasure, Now Closely Guarded,” The Information. ↩︎
19. Report to Congress on the Transformational Export Areas (TEAs) (Washington, D.C.: Export-Import Bank of the United States, December 20, 2023). ↩︎
20. Robert Delaney, “US Export-Import Bank Seeks More Latitude to Counter Belt and Road Initiative,” South China Morning Post, June 28, 2024. ↩︎