Exploring Instability Risks in the U.S.-China AI Rivalry

Why It Matters

If unchecked, the AI race could spill into cyber or military conflict, threatening global security and disrupting technology markets. Understanding the triggers helps governments craft safeguards before escalation becomes inevitable.

Key Takeaways

• AI race could trigger preventive cyber attacks between US, China
• Breakwater simulates strategic decisions under AI development uncertainty
• Beliefs about AGI timeline influence risk tolerance
• Early results identify escalation triggers and stabilizing mechanisms
• Policymakers need frameworks to manage AI rivalry risks

Pulse Analysis

The United States and China have turned artificial intelligence into a centerpiece of their geopolitical competition, pouring billions of dollars into research labs, talent pipelines, and next‑generation computing infrastructure. This race is not merely about commercial advantage; it carries profound national‑security implications as each side seeks to secure a strategic edge that could dictate future military capabilities. Analysts warn that the drive for supremacy may lower the threshold for pre‑emptive actions, especially if one nation believes the other is on the cusp of achieving artificial general intelligence, a milestone that could reshape global power balances.

To probe these dynamics, RAND introduced Breakwater, a tabletop‑style simulation that places two players in the roles of rival AI powers. The game forces participants to make decisions about investment, secrecy, and offensive measures under uncertain timelines for AI breakthroughs. Early sessions reveal that aggressive postures emerge when players perceive a rapid acceleration in the opponent’s capabilities or when trust erodes due to ambiguous signaling. Conversely, mechanisms such as transparent research sharing, joint safety standards, and mutual verification can act as stabilizers, reducing the incentive for pre‑emptive strikes.

The insights from Breakwater underscore a pressing need for coordinated policy frameworks that address both the technical and strategic dimensions of AI development. Governments should consider establishing confidence‑building measures, joint monitoring protocols, and crisis‑communication channels to mitigate misperceptions. For industry leaders, the research highlights the importance of aligning commercial roadmaps with broader security considerations, ensuring that rapid innovation does not outpace the safeguards needed to prevent an AI‑driven escalation spiral. By integrating these lessons, stakeholders can help steer the AI rivalry toward constructive competition rather than destructive conflict.