The AI Competition Will Be Won in the Intelligence Layer

Recent bipartisan efforts on the U.S. Tech PATH Act have introduced a larger question: How can the United States and its allies make trusted digital infrastructure available before other technology defaults harden into place?

Whatever path the legislation takes, the issue it surfaces is central to the next phase of strategic competition. The bill proposes, among other things, establishing a State Department-led effort to help allies and partners procure trusted U.S. cyber and digital technology, with financial and technical support that enables and encourages long-term technology procurement partnerships.

This program has broad national security implications.

Artificial intelligence is becoming the interface between people and data. Increasingly, AI systems will help leaders make mission-critical decisions and inform the sources and information that provide their contextual understanding of high-stakes situations. Whoever shapes that interface shapes the decision environment.

That is why the spread of Chinese model ecosystems deserves serious attention. The risk is not that every Chinese-trained or Chinese-governed system is a covert weapon. However, if these systems are governed by Chinese interests at a structural level, then those same interests implicitly become the frame through which institutions see their own data. Even a small influence, at this scale, has sweeping implications. The AI stack becomes trusted as ground truth – and it does so without this contract of trust being overtly evident to the user.

Technology strategy fails when it stops at the prototype. I learned this directly in my time in government, both at AFVentures and later at the Department of Defense Office of Strategic Capital. Identifying promising technology was never the hard part; building the financing, procurement paths, authorities, and partnerships to convert capability into real use was. The same lesson now applies internationally.

Allied AI infrastructure is ultimately a deployment challenge. Nations need practical ways to adopt secure systems, govern their own data, inspect decisions, and connect with trusted partners without surrendering control of their own digital foundations. That requires something other than siloed innovation. Procurement pathways, technical support, financing mechanisms, and long-term institutional partnerships are an imperative part of that machinery.

China increasingly understands this dynamic. In many regions, firms such as Huawei offer governments an integrated package: telecommunications infrastructure, cloud services, AI capability, technical training, financing, and long-term support. Western alternatives are often stronger in individual technologies, but fragmented across vendors, financing channels, and procurement systems. A government modernizing its digital infrastructure is not evaluating models or hardware in isolation. It is evaluating timelines, sustainment, integration support, and operational reliability.

The American offer, therefore, cannot simply be a better catalog of technologies. It has to be a better system for adoption.

Data sovereignty is not only a matter of privacy or compliance. At the national security level, it means maintaining control over how data is stored, how models are trained and updated, how outputs are generated, and who can audit the underlying decision processes. It means dictating the ground truth. A country that cannot govern those mechanisms itself does not fully control the intelligence systems informing its decisions.

Providing this architecture requires a wholly different model than the Huawei's of the world. Rather than one behemoth cloud-based model that speaks into every mission, national security-minded leaders need specialized models that keep the data close to the user, run of trusted devices and sovereign networks that coordinate when needed, rather than as a default, and only do so through shared standards and provenance. That architecture is more resilient operationally – and it also allows allies to retain control of their own data and decision-making while securely coordinating with partners.

This is how the United States can compete in AI without replicating China's centralized model. The ultimate goal is not to replace one technological dependency with another, but to build architectures that allow allies to retain sovereign control while still operating together securely and effectively.

This is the broader significance behind current debates over trusted technology procurement. The challenge is not whether the United States can innovate. It can. The challenge is whether the United States and its allies can evaluate technology for its capabilities as well as its security and sovereignty before competing systems become entrenched.

It’s promising to see bipartisan cooperation on these high-stakes technology and AI questions. As we move forward, they should remain focused on what will ultimately decide the next phase of strategic competition. It will not be determined solely by who can build the most advanced models – but by which systems nations can trust enough to integrate into the institutions where consequential decisions are made.

Jason Rathje, Ph.D. is the President of webAI Public Sector; former founding Director, Department of Defense Office of Strategic Capital; co-founder, U.S. Air Force AFVentures.

image: U.S. service members and civilian contractors track the Freedom-class variant littoral combat ship USS Wichita (LCS 13) during the annual Fleet Experimentation (FLEX) 2026 event in Key West, Florida, April 26, 2026. Hosted by U.S. Naval Forces Southern Command/U.S. 4th Fleet April 24-30, the exercise integrates commercially developed unmanned systems and artificial intelligence with traditional manned naval platforms. (U.S. Navy photo by Mass Communication Specialist 3rd Class Jasmin L. Aquino)