A Costly Illusion of Control: No Winners, Many Losers in U.S.-China AI Race
In recent years, the rapid development of generative artificial intelligence (GenAI) has intensified the already tense competition between the United States and China in technology. AI has emerged as the focal point of U.S. efforts, primarily aimed at slowing down the advanced model development of Chinese companies. This competitive dynamic has global implications, affecting supply chains and the geopolitical landscape surrounding AI technology.
The Geopolitics of AI
The geopolitics of AI has become a battleground for the United States and China, creating negative externalities that threaten progress towards a secure international framework for AI development. The discussion surrounding artificial general intelligence (AGI) and artificial superintelligence (ASI) has shifted from theoretical concepts to achievable outcomes within a significantly shorter timeframe than previously anticipated.
This article examines the trajectory of U.S.-China AI competition, the risks of unchecked escalation, and potential pathways out of the current zero-sum, winner-takes-all paradigm that has taken root in Washington.
AI Dominance in Technology Competition
Several key factors have led to the current situation where the U.S. and China are locked in a fierce competition for AI dominance. Many U.S. policymakers and think tanks view this competition as zero-sum; the narrative posits that whoever achieves AGI first will gain a significant strategic advantage. This perspective carries profound implications for global AI development, given that U.S. and Chinese companies are the leaders in creating frontier AI models.
The U.S. export control measures targeting AI development are rooted in two main sources:
- The concept of compute governance, which advocates for restricting compute hardware to control AI development, particularly aimed at China.
- The adoption of choke point technologies by the U.S. policy-making community, particularly under the Biden administration, to support measures targeting China.
Key Events and Legislative Actions
Significant regulatory actions have taken place, including:
- In April 2025, the U.S. banned the export of Nvidia’s H20 GPU, a crucial component for running AI applications.
- The implementation of the AI Diffusion Framework aimed at preventing Chinese firms from accessing restricted hardware globally.
These actions reflect a broader strategy within the U.S. government to maintain technological supremacy over China while simultaneously restricting its advancements in AI.
The Cost of Compute Governance
The approach of compute governance has not only disrupted the global AI semiconductor industry but also posed significant risks of conflict. As U.S. regulatory measures have intensified, the effects have rippled through companies across the AI stack, leading to substantial financial losses and limiting innovation.
China’s responses to U.S. controls have included bans on critical minerals, raising production costs and complicating supply chains internationally. This retaliatory strategy further complicates the geopolitical landscape, especially concerning Taiwan, a hub for advanced AI hardware production.
Ignoring the Benefits of AI
While the U.S. policy primarily focuses on the potential downsides of AGI and ASI, it overlooks the numerous benefits that advanced AI could offer in sectors such as healthcare, climate science, and green technology. The justification for U.S. controls often rests on misconceptions around military applications, despite the civilian focus of Chinese AI companies.
International Collaboration and Governance
The unilateral U.S. approach to AI governance has hindered the establishment of a global framework for AI safety and security. The absence of Chinese involvement in these discussions poses severe risks for international AI governance, particularly in mitigating existential risks associated with advanced AI.
Future Directions: Building Guardrails
To address these challenges, several key strategies should be considered:
- Establish a bilateral mechanism between the U.S. and China to discuss AI development and AGI, fostering trust and collaboration on transnational issues.
- Revitalize the Bletchley Park Process to facilitate global AI governance discussions involving major players, ensuring comprehensive participation.
- Promote open sourcing of AI models to enhance transparency and reduce uncertainties in the AGI timeline.
- Include Chinese firms in critical industry forums to build trust and shared responsibility among key technology leaders.
- Launch a major educational campaign to broaden awareness of the risks posed by U.S.-China AI competition and advocate for a unified global AI safety framework.
As the race to AGI accelerates, it is crucial for both nations to recognize the potential for collaboration rather than conflict. The current trajectory must be recalibrated to avoid irreversible damage to international relations and ensure a stable future for AI development.
