US President Donald Trump has rejected calls to slow down artificial intelligence (AI) development, asserting America’s lead in the race, even as Chinese President Xi Jinping advocates for joint human control over the technology. This divergence underscores the global debate on AI regulation and the intensifying competition between the two superpowers.
During a recent visit to the United States, President Xi stated that both nations share a common responsibility to develop and manage AI for the benefit of humanity, acknowledging their roles as leaders in AI advancement. His remarks came as UN Secretary-General António Guterres urged member states at the UN General Assembly to address the uncontrolled development of AI due to its potential risks.
However, President Trump challenged the Secretary-General’s plea, stating that the US would not participate in any global conspiracy to halt AI development. “America is not afraid of the consequences this will bring,” Trump said, suggesting the technology be renamed “Super Intelligence” and emphasizing that the US leads this race, making global efforts to slow its progress unacceptable. This stance aligns with concerns raised by figures like Dario Amodei, CEO of US AI company Anthropic, who previously suggested slowing AI development and preventing China from taking the lead. China, for its part, has warned that “confrontation” and “malicious competition” would undermine global AI governance.
Computing Power and Infrastructure
The capacity and speed to process data and deliver results, known as computing power, are crucial for advanced AI. The US holds a significant lead in this area, according to Epoch AI, an organization researching AI. China’s share of global computing power is estimated at only 14 percent.
The US also leads in advanced AI chips. American company Nvidia designs 60 percent of the world’s AI computing chips, while China’s Huawei produces a comparatively small amount, a Stanford University study found. Beyond chips, the US boasts extensive infrastructure, with 5,400 data centers, reportedly housing ten times more computers than other nations. It has 84 AI-focused data centers, a number that surpasses the combined total of eight other countries. This dominance in infrastructure, technology, and chip production positions the US ahead of China in computing power.
AI Models and User Adoption
While American models like OpenAI’s ChatGPT and Anthropic’s Claude currently dominate, Chinese models such as Deepseek, Alibaba, Z.AI, and Tencent are rapidly gaining popularity. By March 2026, the performance gap between top US and Chinese AI models had narrowed significantly to just 2.7 percent, according to the Stanford AI Index 2026. For instance, the difference between Anthropic’s Claude Opus 4.6 and China’s ByteDance’s Doula Chat 2.0 Preview was only 39 points. Open Router, an AI evaluation platform, indicates that Chinese AI models are ahead in terms of usage, with Deepseek, Z.AI, and Tencent models ranking among the top three. This is attributed to their affordability and ease of use for developers, fostering user confidence and familiarity.
Investment and Talent
The US significantly outspends China in AI development. Investment company Goldman Sachs estimates that $764 billion will be invested in data centers and cloud computing in the US by 2026, involving major companies like Amazon, Microsoft, Google, Meta, and Oracle. In contrast, Chinese companies are projected to spend $102 billion by 2026, although their expenditure on AI model development is rapidly increasing, with an estimated 80 percent rise in 2026, according to TrendForce. This disparity suggests that US companies have greater resources for investing in chips, data centers, and skilled personnel.
In terms of research and talent, China is actively developing a skilled technical workforce. By 2024, China accounted for 27 percent of AI publications (based on English-language materials), compared to 12 percent for the US. While 47 percent of the top 20 AI researchers studied in China, 72 percent of those Chinese-trained researchers are now working in the US, an analysis by Marcopolo shows.
Global Governance and Future Outlook
The United Nations has been actively involved in the debate over AI governance. An independent international scientific panel of the UN issued a preliminary report on July 1, 2026, warning that the window for establishing effective global governance for AI is short. A global dialogue on AI governance, organized by the UN, commenced in Geneva on July 6, 2026. China is also set to host the World AI Conference 2026 and a high-level meeting on global AI governance in Shanghai in July 2026.
Both the US and China are reportedly discussing mechanisms for regular dialogue on potential AI risks and national security challenges. The UN will continue its efforts to bring nations together to adopt common principles and commitments for AI governance. Upcoming events like Black Hat Europe 2026 on December 8 and SANS AI Cybersecurity Summit Fall 2026 on November 2-3 will further explore AI security and regulatory landscapes. However, the detailed outcomes of the recent Trump-Xi summit regarding AI, the implementation plan for renaming AI to “Super Intelligence,” and the final stages of UN Secretary-General Guterres’s successor selection process remain to be fully disclosed.
