Huawei Ascend and DeepSeek signal China is winning the AI cold war

For more than two decades on the Army Staff, part of my job was to recommend which nations received American weapons, training, and doctrine and which did not. The choice rarely came down to which weapon system performed best at a range. It came down to alliance.

A country that trained on American equipment, spoke our tactical language, and built its systems around our supply chains remained beholden to Washington for a generation. One who returned to Moscow or Beijing was dragged into someone else’s orbit.

This lesson has stayed with me. Great powers rarely prevail because they have the best weapons; They prevail because other nations choose to build their militaries, economies, and ultimately their futures around their systems. Washington risks forgetting this lesson in the race to build the world’s dominant computing platform.

ROBERT MAGINNIS: TRUMP’S ANKARA STATEMENTS REVEAL A MAJOR STRATEGY HIDING IN APPEARANCE

It’s not just a software race

While Washington debates which chatbot makes the clearer draft of an article, Beijing is building something much more ambitious than a single flagship model. This distinction separates today’s technology race from tomorrow’s world order. Consider that China’s Huawei is preparing to double production of its Ascend processors to 1.6 million chips in 2026, and that Chinese developers at DeepSeek are already tuning their latest models to run specifically on this Huawei silicon.

YOU TODD YOUNG: THE HIDDEN DANGER THAT CHINESE SHIPS MAY BRING TO OUR SHORES

Meanwhile, congressional hearings and cable news segments continue to ask which model scored highest in the latest benchmark; An interesting question but not decisive.

The great wars of history were won not by a single best weapon, but by nations that could produce energy, build factories, and produce the industrial output needed to prevail.

CHINA’S ROBOT-OPERATED HOTEL WILL OPEN TO THE PUBLIC IN 2027

This technology is no different.

America considers this competition as a technology race. China sees this as a civilization-building project.

The mistake Washington continues to make is to assume that there is a single technology race. No. There are multiple competitions occurring simultaneously, each reinforcing the other.

AI Power Stack

National competitiveness in this competition is based on what I call the AI ​​Power Stack: interdependent layers that together determine technological power. It all starts with abundant, reliable electricity. Each of the newest data centers can draw more than a gigawatt, roughly the output of a nuclear reactor, and China now produces more than twice as much electrical power as the United States; can route the centralized system to computing clusters much more easily than our fragmented grid allows.

Above energy are semiconductors and steel mills of the digital age. America’s effort to rebuild domestic chip manufacturing is real progress.

On top of the chips lies the computing infrastructure: data centers, networking, and cooling that turn processors into usable capability. Only after these foundations are formed, the models built on them become decisive, and the models are almost the issues that Washington discusses.

There are applications on the models: factories, hospitals, farms and command centers where this technology is used. And above each layer is the layer that Washington discusses least: the ecosystem of developers, companies, universities, investors and allied countries that decide which technology will become the world standard.

The great wars of history were won not by a single best weapon, but by nations that could produce energy, build factories, and produce the industrial output needed to prevail.

Beijing understands the ecosystem

China isn’t just trying to invent the technology of tomorrow. It tries to be the platform on which tomorrow’s technology will work.

Beijing understands this upper layer better than Washington realizes. Rather than chasing a single groundbreaking model, Chinese companies are pricing their models aggressively and believing that adoption will increase over time.

As of February 2026, Chinese open source models were attracting more weekly token traffic than American models in the world’s largest model market, with four of the five most used systems worldwide being built in China. One venture capital partner estimates that most American startups are now built on China-based models because they are cheaper to run.

Banks in Singapore, telecom operators in Indonesia and government platforms in Malaysia are already working on Chinese models and Huawei hardware. History shows that the technologies that reshape civilization are rarely the ones that engineers admire most. These are the products that businesses, governments, and consumers have adopted and continue to improve for decades.

Why does adoption trump grace?

This model is not new. The Internet won out not because it is the most secure network ever designed. It prevailed because millions of people built on it. Cloud computing has reshaped global commerce for the same reason. This technology will follow the same path.

This reality explains why today’s debate over explicit and implicit models is much more than a technical disagreement. Closed models emphasize security, control, and carefully managed distribution. Open models enable universities, startups, and allied nations to develop new applications and accelerate their adoption.

The debate is not just about protecting intellectual property. It’s about determining which technology ecosystem the next billion users and the nations in which they live will trust. This larger strategic competition is one that I explore in more detail in my book. “The New AI Cold War: Freedom and Versus Tyranny in the Age of Machine Empires.”

America still has the better hand

Washington is not standing still. President Donald Trump’s AI Action Plan, released in July 2025, directs the Departments of Commerce and State to put together full-scale American export packages of assembled hardware, models, software, and standards for allies and partners abroad. This is exactly the right instinct.

National competitiveness in this competition is based on what I call the AI ​​Power Stack: interdependent layers that together determine technological power. It all starts with abundant, reliable electricity.

This strategy recognizes the fundamental truth of this race: America cannot export chips alone; It needs to export its entire technology ecosystem.

It approaches this technology the way effective security cooperation treats weapons systems: not a single sale, but a decades-long relationship.

CLICK FOR OTHER OPINIONS OF FOX NEWS

But the plan will only succeed if Washington backs it with the same urgency as defense budgets and semiconductor legislation: faster permitting for power generation and transmission, continued investment in domestic chip production, and a willingness to compete on price and not just talent in emerging markets that China is actively courting.

In conclusion

CLICK TO DOWNLOAD FOX NEWS APPLICATION

The nation that wins the AI ​​Cold War won’t necessarily create the smartest chatbot. It will build the ecosystem that the rest of the world chooses to trust, adopt and expand. History shows that once these ecosystems take root, they shape alliances, trade, military power, and political influence for generations.

Beijing now appears to understand this fact better than Washington. America still has the strategic advantage, but only if it recognizes the real battlefield before decisive campaigns begin.

CLICK HERE FOR MORE FROM ROBERT MAGINNIS