The U.S.-China AI Race: Collaboration Amidst Competition

The artificial intelligence (AI) landscape is characterized by intense rivalry between the United States and China, yet a surprising degree of collaboration persists. An analysis of over 5,000 research papers presented at the Neural Information Processing Systems (NeurIPS) conference revealed that approximately 3% involved joint authorship from U.S. and Chinese institutions. This collaboration has remained stable, with 134 out of 4,497 papers in 2024 reflecting similar partnerships. Notably, significant models like Google's transformer architecture and Alibaba's Qwen have been co-opted across both nations, indicating a shared ecosystem that benefits both sides.

The US and China Are Collaborating More Closely on AI Than You Think

In the U.S., the legislative landscape is evolving in response to concerns over national security and technological dominance. The "AI Overwatch Act," introduced by Representative Brian Mast, aims to grant Congress authority over AI chip exports, particularly to China. This bill follows President Donald Trump's approval for Nvidia to sell its H200 AI chips to Chinese firms, a move that has sparked controversy. Critics, including Anthropic CEO Dario Amodei, have likened the decision to selling nuclear weapons to North Korea, emphasizing the potential risks to U.S. security. The bill reflects a broader institutional struggle over tech policy, with differing views on whether to leverage U.S. technological advantages or restrict access to maintain a competitive edge.

Anthropic CEO Says AI Chip Sales to China Like Selling Nukes to North Korea

A recent Ipsos and Google poll indicates that global enthusiasm for AI significantly outpaces that in the U.S. Approximately 40% of American adults reported using generative AI in the past year, compared to nearly 75% in Brazil and 85% in countries like Nigeria and India. This disparity highlights a broader concern in the U.S. about the economic implications of AI, with many citizens expressing apprehension about job displacement. In contrast, individuals in emerging markets view AI as a tool for economic advancement, suggesting a divergence in perceptions of AI's potential benefits.

Much of the world seems more enthusiastic about AI than Americans, a poll finds

China's approach to AI is marked by a focus on integrating AI into everyday life, with ambitions to become the world's first AI society by 2035. The Chinese government is actively promoting AI as a public utility, embedding it across various sectors, including agriculture and healthcare. This strategy is supported by a robust ecosystem of over 6,000 AI companies, with significant investments in AI chip production and open-source models. The recent rise of models like DeepSeek has galvanized the Chinese AI sector, demonstrating the country's capacity for rapid innovation despite U.S. export restrictions.

China’s AI industry output value estimated to exceed 1.2 trillion yuan in 2025: senior officialWhy China’s AI models will have greater global appealUS Rules Create Chip Bottleneck for China’s AI Push

Critics of the U.S. approach argue that overly restrictive export controls may inadvertently strengthen China's AI capabilities. The Trump administration's recent decision to allow Nvidia to export H200 chips to China has been met with skepticism, with concerns that it could accelerate China's military and technological advancements. Conversely, proponents of the policy argue that maintaining some level of engagement with Chinese firms could prevent them from developing independent alternatives, thereby preserving U.S. technological leadership.

America Can’t Win the AI Race by Retreating

The U.S.-China AI race is not merely a contest of technological supremacy but also a complex interplay of collaboration and competition. As both nations navigate their respective strategies, the implications for global AI development and international relations remain profound. The future will likely hinge on how effectively both countries can balance cooperation in research with the need for national security, while also addressing the broader societal impacts of AI technology.

America Can’t Win the AI Race by Retreating