America vs China: The AI Battle to Own the Future

The world may not have officially entered an artificial intelligence war, but it is already living inside an AI power struggle.

The United States and China are competing to control the technologies that may define economic growth, military strength, scientific discovery and political influence for decades. The weapons in this contest are not only missiles or drones. They are advanced computer chips, data centres, algorithms, research talent, cloud infrastructure, global standards and access to enormous quantities of data. Both countries present their ambitions as beneficial to humanity. Yet their policies increasingly speak the language of winning, leadership and strategic advantage.

America’s AI Action Plan, released in July 2025, explicitly describes AI development as a race the United States must win. Its strategy is built around accelerating innovation, expanding domestic infrastructure and exporting American technology, models and standards to allies. China, meanwhile, is embedding AI across manufacturing, science, education, public services and governance through its “AI Plus” initiative. By 2027, it aims for AI agents and new intelligent devices to reach a penetration rate exceeding 70 percent in key areas.

This is not a simple competition between one country that innovates and another that follows. According to Stanford University’s 2026 AI Index, the United States produced 59 notable AI models in 2025, compared with China’s 35. However, China led in the volume of AI publications, citations and patent grants. By March 2026, the performance difference between the strongest American and Chinese models had narrowed to only 2.7 percent. 

The race is therefore real, close and rapidly intensifying. At its centre lies the semiconductor. Advanced AI requires enormous computing power, and that power depends on specialised chips. Washington has repeatedly restricted China’s access to advanced processors and semiconductor-manufacturing equipment, arguing that such technologies could support military modernisation. Although some licensing policies have changed, chip access remains a tool of strategic pressure.

China has responded by investing in domestic chips, efficient models, industrial applications and alternative technological ecosystems. What began as commercial competition is producing two increasingly separate AI worlds. Countries may eventually be pressured to choose whose chips, clouds, platforms, security systems and technological standards they will use. This is where the Global South enters the battlefield, not necessarily as a combatant, but as a market, data source and zone of influence.

America wants allied countries to build on American AI technology. China presents its approach as more open to international cooperation and has announced training opportunities, AI application centres and capacity-building programmes for developing nations. In July 2026, China also promoted a new World Artificial Intelligence Cooperation Organization and promised 5,000 AI training opportunities for developing countries over five years.

Both approaches offer genuine opportunities. Developing countries need computing access, technical training, investment and affordable digital infrastructure. But assistance can also create dependence. A country that relies entirely on foreign AI systems may surrender control over its data, public services, information environment and national decision-making. The most dangerous part of the contest is its military dimension.

The United States is expanding AI use across its national-security institutions and armed forces. China’s military strategy has been described by the US Department of Defense as moving towards “intelligentized warfare,” in which AI supports surveillance, command systems, autonomous platforms and battlefield decisions. China publicly maintains that military AI must remain under human control and says it does not seek an AI arms race. These public commitments are important, but declarations are not enforceable safeguards.

AI can analyse satellite images, identify possible targets, guide autonomous drones, generate military plans and conduct cyber operations at speeds humans cannot match. Speed may appear to make defence more effective, but it can also make war less controllable. A flawed system could misidentify a civilian vehicle, interpret routine military movement as an attack or produce a false warning during a nuclear crisis.

When two rival powers fear falling behind, caution begins to look like weakness. Each side accelerates because it assumes the other is accelerating. This is how an innovation race can become an arms race. The threat is not limited to physical weapons. AI-generated propaganda, deepfakes and automated influence campaigns could manufacture speeches, create false evidence of attacks or provoke public anger before facts can be verified. The United Nations has warned that AI-enabled cyberattacks can damage critical infrastructure within minutes, while fabricated audio and video can intensify political tension and even trigger diplomatic crises.

Humanity should not be forced to choose between American technological dominance and Chinese technological dominance. It should demand a third option: internationally governed AI development centred on human survival and dignity.

Life-and-death decisions must remain under meaningful human control. AI should never independently authorise nuclear action or select human targets without accountable human judgement. Countries need binding rules for autonomous weapons, military AI audits, crisis hotlines and mandatory reporting of serious AI failures. The International Committee of the Red Cross recommends prohibiting unpredictable autonomous weapons and systems designed to target people without sufficient human control.

Civilian AI also requires cooperation. Scientific research, medical discoveries and climate technologies should not become inaccessible simply because they emerge on the wrong side of a geopolitical divide. Developing countries need a genuine place in global rule-making, not invitations to choose a technological master.

America and China may believe they are competing to own the future. But the future does not belong exclusively to either of them. It belongs to the billions of people whose jobs, security, privacy, education and understanding of truth will be transformed by decisions they never helped make. The real victory will not be producing the most powerful machine. It will be ensuring that, when the race ends, humanity is still in control.