“The real danger isn’t that China is faster... it’s that the U.S. is more divided.”

Introduction: The Illusion of Technological Superiority

In the geopolitical arena of the 21st century, artificial intelligence (AI) is no longer simply a military advantage: it has become the battlefield itself. From automatic target detection to the autonomous orchestration of precision strikes, cognitive warfare and the domination of logistical networks, military AI is establishing itself as an unprecedented power multiplier (Zequeira, 2024). Yet the United States risks losing this war, not for lack of resources, but for lack of strategic coherence. While China structures its military doctrine around the notion of intelligent warfare (智能化战争), the United States still seems trapped in schemes of technological superiority inherited from another century. Indeed, as journalist Fareed Zakaria observed, the global order has become “post-American.”(Araya & He, 2024). Hence, it is no longer enough to have the best algorithms. We need to integrate AI into military doctrine, organization, chains of command and industrial ecosystems, because even if operations such as “where is daddy” carried out by the Israeli government may call into question the ethical dimension of this project (Human Rights Watch, 2024), tomorrow's war is an AI war, where to protect interests and avoid hegemony, armies must remain competitive, as the war in Ukraine proved. So, in a world where AI is becoming the driving force behind military transformation, can the United States maintain its lead in the face of a strategically unified, technologically aggressive and doctrinally disruptive China?

I. The American Illusion: Believing Technological Superiority Is Enough

The United States will not lose the AI war for lack of technology, but for lack of strategy. Their fundamental mistake is to believe that technological superiority is enough to guarantee military domination. However, in an algorithmic war, it's not the one who innovates the most who wins, but the one who integrates the fastest.

The example of the U.S.-led Project Venom, where an autonomous F-16 piloted by an AI outperforms a human fighter in a combat simulation, shows that the capabilities are there (Shoup, 2024). But they remain isolated. The United States excels in R&D. Indeed, OpenAI, Nvidia, Anthropic and Google DeepMind dominate the global ecosystem. However, the military transfer of these technologies is hampered by institutional, political and cultural obstacles. The case of Project Maven, aimed at exploiting AI to analyze images captured by drones, is revealing: under pressure from its employees, Google withdrew from the project, citing ethical reasons (Berman, 2025). Microsoft and Amazon, too, are reluctant to commit themselves fully. As a result, US military AI is fragmented, lacking a coherent employment doctrine, and locked into a logic of technological silos.

A second obstacle is the defensive logic of the AI Export Control Framework, which, since 2025, classifies countries into three rigid categories: “Trusted” nations (the U.S. and 18 close allies) have free access to AI chips but face restrictions on cloud provider deployment; “Intermediate” countries, such as Israel or Poland, longstanding allies, face strict quotas and cloud deployment limitations, even when using American firms; and “Restricted” countries like China, Russia, Iran, and North Korea are completely cut off from advanced semiconductors (Kennedy, 2025). This framework, intended to protect U.S. supremacy, paradoxically weakens its ecosystem: it limits U.S. tech expansion, slows down allied collaboration, and drives regional powers toward China.

By trying to lock down computing power, Washington has inadvertently outsourced algorithmic influence to Beijing, which imposes no access conditions. Instead, China embeds its technology in a sovereign, aggressive, and open infrastructure strategy. America invents; China deploys. Until the U.S. sheds its outdated superiority complex and transforms its tech leadership into an integrated military architecture, it will remain vulnerable in a world where strategic speed is the primary weapon.

II. China Thinks Like a Nation at War

The U.S.’s challenge with China is not raw power, it is strategy. Beijing sees AI not as a tool, but as a total war architecture, while Washington still sees it as a modular capability. This divergence is reshaping the global balance of power. Unlike the U.S., China recognizes no separation between civilian tech and military doctrine (Stokes, 2024). Its policy of civil-military fusion (军民融合) turns every private tech advancement into a strategic asset. Companies like Baidu, Tencent, and iFlyTek are not autonomous, they are integrated into the People Liberation Army’s command and innovation structure (Stokes, 2024). By 2027, China’s top 2,000 firms will be required to allocate over 50% of core IT spending to AI (Xie et al., 2023).

Since 2017, China has embedded AI into its military modernization plan, targeting an “intelligentized” force by 2035 (Stokes, 2024). Central to this is “systems destruction warfare”: not defeating the enemy on the battlefield, but dismantling their networks such as satellites, command centers, communications (Stokes, 2024). The PLA envisions AI that designs entire operational plans, effectively replacing human staff officers with a “command brain” (Stokes, 2024).

AI, in this context, is a cognitive weapon: it disrupts adversary C3I (command, control, communication, information), generates targeted deepfakes, orchestrates offensive cyber operations, and leverages “adversarial AI” to undermine enemy systems. Meanwhile, the U.S. operates in silos. The Air Force tests AI drones (Shoup, 2024), the Pentagon partners with Palantir, but no unified doctrine links these efforts. Worse, America’s strategy is reactionary and based on building “chokepoint” tactics like restricting A100 chips and export controls. These measures open a strategic void, quickly filled by China’s open-source models, subsidized cloud, and turnkey digital infrastructure for non-aligned nations (Wang, 2025).

China is not chasing superiority, it is building coherence. Every innovation is a power vector. Until the U.S. adopts a global military systems mindset for AI, it won’t lose because China is faster, but because China is more unified.

III. What Must Be Done: Unify Civilian Tech and Military Doctrine

The real challenge for the United States is neither technological nor budgetary. It is doctrinal and structural. Faced with a China that sees artificial intelligence as a total war system, America continues to treat it as a technological tool to be added to its arsenal. This approach is now obsolete. AI is not a military accessory: it redefines the very mechanisms of power. If the United States wants to remain a world military force, it must make AI the backbone of its defense strategy, in the same way as nuclear power was during the Cold War (Zeng, 2025). 

This presupposes, first and foremost, the creation of an Integrated Algorithmic Command, joint and civil-military, with an operational, industrial and doctrinal mandate. This algorithmic center of gravity would not be a coordination agency, but a command structure with real authority over on-board cognitive systems, tactical LLMs, on-board AI models in autonomous fleets, and the synchronization of the C2I (command-control-information) chain. The next step is to break with the passive outsourcing posture towards Big Tech. Google, Microsoft, Palantir and Nvidia can no longer be treated as technology subcontractors. They must become sovereign defense partners, legally and strategically integrated into the military architecture. This implies securing their participation via an AI Defense Acceleration Act, which guarantees their commitment, compensates for their commercial exposure, and accelerates the transfer of prototypes to the field. 

On the external front, there is an urgent need to break out of the bureaucratic straitjacket of the AI Export Framework, which, by classifying allied nations as “intermediaries”, weakens the American strategic ecosystem (Wang, 2025). We need a NATO+ Algorithmic Alliance, with its own military LLMs, distributed sovereign clouds, algorithmic interoperability standards and shared doctrine. The deterrent of the 21st century will not be nuclear, but cognitive, distributed, predictive and interoperable. This posture does not consist in over-militarizing AI, but in regaining the strategic initiative in the face of a China that is already imposing its architectures, standards and logic on the global digital environment. It's a paradigm shift: AI is no longer a field of competition, it's the battlefield itself.

Conclusion

The United States must recognize that AI is not just another emerging technology, it is a total strategic system. It redefines intelligence, logistics, planning, coordination, and deterrence. In this environment, decision speed is the ultimate weapon. The actor with the most agile, interconnected, and adaptive AI system will dominate not by firing more missiles, but by paralyzing the opponent’s ability to act.

To stay ahead, the U.S. must act like a nation at war: mobilize industry, fuse doctrines, standardize technologies, and train officers in algorithmic thinking. The future of American power won’t be decided in Pentagon hangars but in Nvidia’s GPUs, military-grade LLMs, AWS data centers, and the strategic doctrine built around them.