U.S. policy is often reported through announcements, personalities, and regulatory skirmishes. Far less attention is paid to the economic mechanisms that actually move structures and determine outcomes.
To understand how the White House is organizing a multi-pronged strategy for AI adoption and export, and how its pieces are meant to work together in practice, I had an exclusive sit down with Michael Kratsios, assistant to the president and director of the White House Office of Science and Technology Policy.
The fundamental issue you speak about at the summit is the widening AI adoption gap between the developed and developing world. What makes that a concern for the White House right now?
The divergence in AI adoption between developed and developing countries is growing every day. We see the world in two broad categories, and different tools are needed for each.
Developing countries are at risk of falling behind at a fundamental inflection point. That is why we urge them to prioritize AI adoption in sectors that deliver concrete benefits: healthcare, education, energy infrastructure, agriculture, and citizen-facing government services.
For too long, countries seeking development support faced a false choice. We believe the American AI Exports Program offers a different path: trusted best-in-class technology, financing to overcome adoption barriers, and deployment support so governments can learn how and where to use these tools.
America remains the undisputed leader in AI, from GPUs to data centers to frontier models and applications. That leadership brings with it a responsibility to share the foundations of a new era of innovation. We stand ready to work with partners around the world so creativity, freedom, and prosperity shape today’s technological revolution.
A lot of governments say they want AI leadership. Your delegation came in talking about real AI sovereignty, rejecting global governance, and launching an export program with multiple prongs. What is fundamentally different about this approach, and how should countries understand the system you’re building?
The hope of the United States is that the pursuit of real AI sovereignty, the adoption and deployment of sovereign infrastructure, sovereign data, sovereign models, and sovereign policies within national borders and under national control, will become an occasion for bilateral diplomacy, international development, and global economic dynamism. The American AI Exports Program exists to make that happen.
Real AI sovereignty means owning and using best-in-class technology for the benefit of your people, and charting your national destiny in the midst of global transformations. We urge nations to focus on strategic autonomy alongside rapid AI adoption rather than aiming for full self-sufficiency. AI adoption cannot lead to a brighter future if it is subject to bureaucracies and centralized control.
We deeply believe that the best pathway for the developing world to fully realize the untold benefits of AI is through the adoption of the American AI stack. The American AI stack has the best chips, the best models, and the best applications in the world, and that is what countries ultimately need to deploy AI effectively.
When you say the American AI stack, are you talking about selling products, or shaping the foundation on which countries build while keeping sensitive data under national control?
Working with the American AI stack allows nations to build on the best technologies in the world while keeping sensitive data within their borders. Independent partners are critical to unlocking the prosperity AI adoption can deliver. That is why the President launched the American AI Exports Program.
American companies can build large, independent AI infrastructure with secure and robust supply chains that minimize backdoor risk. They build it, and it belongs to the country deploying it.
If this is an adoption strategy, then cost and complexity become the bottlenecks. Your public remarks emphasize financing and deployment sophistication as the two biggest hurdles for developing countries. How are you actually removing those barriers?
Developing countries face two major obstacles to AI adoption. One is financing. The AI stack is expensive. Through the energy and material demands of its infrastructure, it brings the digital transformation of our world back into physical reality. Data centers, semiconductors, power production all require real labor and real resources.
The second barrier is a deficit in the technical sophistication needed to deploy AI tools effectively. To address this, we announced a U.S. government-wide suite of support initiatives to facilitate global adoption of trusted AI systems, create a competitive and interoperable AI ecosystem, and advance the American AI Exports Program in both developed and developing partner nations.
Spell out that suite. What are the prongs, capital, integration, standards, execution, and which agencies are being activated?
We unveiled a new set of initiatives across the federal government supporting the American AI Exports Program, which was launched by executive order last July.
The first new initiative within it is the National Champions Initiative. It is designed to include the leading technology companies of partner countries directly into the American AI stack. We want the best technologies from all our partners and allies to be part of that ecosystem wherever the American AI stack goes.
The second is a full suite of financing and funding opportunities. We are mobilizing support through the U.S. International Development Finance Corporation, the Export Import Bank, the Millennium Challenge Corporation, the U.S. Trade and Development Agency, and a new World Bank fund, with additional programs launched by Treasury and other parts of the U.S. government. The message is simple: this is serious. Every possible financing avenue is being brought to bear.
The third is the creation of the U.S. Tech Corps. It is a reimagining of how the Peace Corps can make an impact in the modern era. We are seeking Americans with technical backgrounds who can help deploy American technology abroad, because there is no better tool to drive economic development, health improvements, and quality of life gains than AI.
And finally, we believe one of the fastest ways to drive global adoption is through standards, particularly as the next wave of innovation centers on AI agents. How those agents communicate and coordinate their actions will benefit from unified standards, which is why NIST has launched a dedicated initiative.
The National Champions Initiative is easy to misunderstand. Critics hear American stack and assume dependency. Your framing suggests the opposite, integrating partner champions so countries do not have to choose between importing the stack and building domestic capability. Is that the point?
Exactly. To integrate partner nation companies with the American AI stack and ensure that no country has to choose between completing the stack and developing domestic AI, we established the National Champions Initiative. Partners need the opportunity to build native technology industries, and facilitating that is a core part of the exports program.
You have also criticized previous U.S. approaches to AI diffusion for restricting partners. What did that get wrong strategically?
The previous approach treated partners as second-tier actors with significant restrictions on access to advanced technology. That was a lose-lose AI diplomacy strategy. It cut off partners from the best technology and limited American companies from competing globally.
Under President Trump, the United States is rethinking how it advances international development and how technology can deliver lasting impact. We believe both developed and developing countries can build sovereign AI capability if given the chance.
Let’s talk about the Tech Corps, because it would be easy to dsmiss it as a feel-good addition. In your model, it sounds like an execution layer. What would these teams actually do on the ground?
These will be like Peace Corps volunteers, except the focus is on technology. We are looking for people with technical backgrounds who want to help implement AI solutions.
If a country wants to improve agriculture through precision farming, apply AI to healthcare systems to improve hospital efficiency, or modernize digital public services, American technologists through the Tech Corps and the Peace Corps will be able to support those efforts.
A lot of young people today care deeply about real-world impact. What is special about this moment is that the United States has incredible technology, the best chips, models, and applications, and we are being more deliberate about sharing it.
You put unusual emphasis on AI agents and interoperability. Why does the White House see standards as a strategic lever now?
The next wave of AI innovation over the next year or two will center on agents. How those agents communicate and orchestrate their actions would benefit greatly from unified standards. NIST has launched an initiative to develop standards for agents so these systems can interoperate securely and effectively.
You also linked this export architecture to supply chains, from chips to data centers to power and minerals. Where does Pax Silica fit? Is it the hard backbone complement to the adoption layer?
Pax Silica is a broader alliance focused on supply chain challenges that the United States and many partner nations have faced. It is a small, select group of countries working together to alleviate these challenges. India is a tremendous addition.
AI adoption depends on secure physical inputs. The AI stack is tangible: data centers, semiconductors, power generation. Pax Silica helps address those vulnerabilities while the exports program accelerates adoption. They are complementary.
Since India hosted the summit and joined Pax Silica, what role do you see for India within this strategy?
India is a technology powerhouse. It graduates an incredible number of engineers, has deep domestic talent, and is building strong products and applications. We look forward to working with them.
India has long been a strong partner in how the United States shares technology abroad. Our major hyperscalers have data centers and research operations here and employ large numbers of Indian engineers. We believe many Indian companies can ultimately become part of the American AI stack.
When critics frame this as being about China, you resist that characterization. How does the administration view competition?
We do not see this as being about any one competitor. This is about the fact that the United States has the best AI technology in the world, and many countries want it in their ecosystems. We are excited to share it and build mutually beneficial partnerships globally.