Anthropic is releasing a new policy paper warning that China could eventually overtake the United States in the global AI race if Washington fails to strengthen chip restrictions and defend its technological advantage.

The paper argues that advanced artificial intelligence will soon become one of the world’s most important geopolitical tools, shaping military power, cyber capabilities, economic growth, and global influence. Anthropic said the US currently leads in the technologies needed to build frontier AI systems, but warned that the gap with China may close faster than many policymakers expect.

The company described the coming years as a critical window for the United States and its allies. According to Anthropic, decisions made before 2028 could determine whether democratic nations or authoritarian governments shape the future rules and standards surrounding advanced AI.

Chips shape AI leadership

Anthropic’s paper places advanced semiconductors at the center of the global AI competition. The company said powerful AI systems require enormous computing resources, most of which still depend on chips designed by American and allied firms.

Much of that advantage comes from companies like NVIDIA, whose AI GPUs currently power a large share of the world’s frontier AI systems. Anthropic’s paper argued that America’s leadership in advanced chips remains one of its strongest strategic advantages in the race against China.

The report argued that US export controls have slowed China’s ability to scale frontier AI development. However, Anthropic claimed Chinese companies continue finding ways to access advanced computing through overseas infrastructure, smuggling operations, and loopholes in current restrictions.

Anthropic also pointed to what it called “distillation attacks,” where AI firms study and replicate the behavior of leading American AI models to speed up development.

The company stressed that China already has world-class engineering talent and significant state-backed investment. In Anthropic’s view, compute access remains the biggest barrier preventing Chinese firms from fully matching top American AI labs.

AI becomes strategic technology

The paper framed AI as more than a commercial technology race. Anthropic warned that future AI systems could reshape national security, military planning, cyberwarfare, and digital surveillance.

The company raised concerns that authoritarian governments may use advanced AI to strengthen censorship systems, automate mass monitoring, and expand state-controlled digital infrastructure into other regions. Anthropic also warned that highly capable AI models could dramatically improve offensive cyber operations by accelerating vulnerability discovery and attack planning.

At the same time, Anthropic argued that the US still holds several major advantages, including stronger chip ecosystems, leading AI companies, and broader access to capital for large-scale AI infrastructure.

The paper suggested that stronger enforcement of semiconductor restrictions could help the United States maintain a lead of up to two years in advanced AI systems by the end of the decade.

Two futures for 2028

Anthropic outlined two possible scenarios for the AI landscape in 2028.

In the first scenario, the US and allied nations maintain a commanding lead in frontier AI development. Anthropic said that the outcome would allow democratic countries to shape global AI standards, safety practices, and infrastructure deployment.

The second scenario paints a more concerning picture. Anthropic warned that weaker enforcement and broader access to advanced chips could allow Chinese firms to close the gap with American AI companies and expand Beijing’s influence over global AI infrastructure.

The company called for tighter export controls, stronger protections against technology theft, and wider adoption of American AI systems across allied markets.