Trump and Xi face off over AI supremacy as US-China rivalry intensifies

Donald Trump has been unambiguous about what matters most to him in the race for artificial intelligence dominance: winning it. "Whoever wins AI, WINS!" he declared, naming China as America's primary competitor without hesitation. Xi Jinping has been more measured in his public statements, but analysts who study Beijing's long-term strategy see little daylight between the two leaders' conviction that AI represents the ultimate prize—a technology that will reshape economic power, military capability, and geopolitical influence for decades to come.

The question of who is actually ahead depends entirely on how you measure the race. American companies—Anthropic, Google, OpenAI—dominate the frontier, producing the highest-performing AI models on most standardized benchmarks. In March 2026, Anthropic's leading model held the top position by a margin of just 2.7 percent over its nearest rival, according to researchers at Stanford University. Yet that narrow gap tells its own story. In February 2025, a Chinese model from DeepSeek briefly matched the best American system, suggesting the distance between the two countries is closing faster than many expected. The US advantage is partly structural: American companies have access to vast pools of capital, with stock markets rewarding AI investment and the broader economy increasingly dependent on the sector's growth. China, meanwhile, faces crippling restrictions on access to advanced microchips—the hardware essential for training cutting-edge models—imposed by Washington as part of its technology containment strategy.

But China is pursuing a fundamentally different strategy, one that may prove equally consequential. Rather than chasing frontier models, Chinese developers have focused on releasing open-weight systems—downloadable, modifiable code that functions like open-source software. Models such as Moonshot AI's Kimi K3 perform nearly as well as American frontier systems while remaining freely available to anyone with the technical capacity to use them. New Chinese models arrive almost monthly, trailing their US counterparts by only a few months on performance benchmarks. The strategic logic is clear: if the US leads at the frontier, China can still gain enormous economic and geopolitical leverage by making capable AI cheap, ubiquitous, and accessible. China also holds structural advantages in energy costs—a critical factor since AI training consumes vast amounts of electricity—and has become the world's largest manufacturer of robots, with more than two million operating in Chinese factories. The country accounts for more than nine-tenths of global humanoid robot shipments, a field many analysts believe will reshape manufacturing and labor worldwide.

Chinese developers have demonstrated what analysts call "remarkable resilience and innovation" despite the chip restrictions. US officials have alleged that Chinese companies closed performance gaps through distillation—a technique allowing advanced models to learn from more capable systems—though Beijing's foreign ministry has framed AI development as driven by a vision of shared benefit and extensive consultation. The gap between the two countries' regulatory philosophies is equally stark. Trump has dismissed AI safety concerns as a hoax and argued that the only guardrails the technology needs are a "high IQ" president. He has rejected proposals for industry-wide slowdowns, fearing they would allow Chinese competitors to catch up. Washington's strategy prioritizes innovation and minimal regulation, betting that American labs can stay ahead through speed and investment. Xi, by contrast, has urged Chinese leaders to keep close watch on AI's risks and ensure the technology remains "under human control." Beijing has introduced guidelines on data usage and algorithms—not legally binding, but setting expectations while preserving room for innovation. The two countries worry about fundamentally different dangers: Washington fears capability escaping human control; Beijing worries about capability destabilizing domestic order.

When Trump and Xi meet on Thursday, AI will be among the most consequential topics on the agenda. Yet cooperation remains uncertain. The two countries have cooperated on existential threats before—nuclear weapons, for instance—but only when sharing a common understanding of the danger. That shared understanding does not exist with AI. Ahead of the summit, US Treasury Secretary Scott Bessent and Chinese Vice Premier He Lifeng discussed establishing a notification mechanism for serious AI incidents, a modest step toward coordination. But Trump's recent messaging on China has grown more antagonistic, and there is real risk that the competitive framing of the AI race will intensify rather than ease. Policy analysts have warned against allowing an "unhealthy version of the race narrative" to perpetuate, one that could escalate an already fierce rivalry into something more destabilizing. What happens in that room will matter not just to Washington and Beijing, but to every country trying to navigate a world where two superpowers are racing toward a technology neither fully understands and neither is certain how to control.