Strategy and Power in an Uncertain AI Future
Jake Sullivan & Tal Feldman | January 27, 2026
In Geopolitics in the Age of Artificial Intelligence, Jake Sullivan and Tal Feldman examine how artificial intelligence is reshaping global power, competition, and national security. Rather than assuming a single future for AI, the article argues that effective strategy must account for uncertainty and multiple possible trajectories of technological development.
AI, Power, and Strategic Uncertainty
The authors note that debates about AI often rest on hidden assumptions. These include whether AI progress will accelerate toward superintelligence or remain uneven, whether breakthroughs will be easy or hard to replicate, and whether China is racing to the technological frontier or prioritizing diffusion. Relying on only one assumption, they argue, risks costly strategic errors.
To address this uncertainty, the article introduces a framework of eight possible “AI worlds.” These scenarios are built around three variables: the pace of AI progress, the ease of catching up to breakthroughs, and China’s strategic choices. This approach encourages policymakers to plan across multiple futures rather than betting on a single outcome.
Sources of AI Power
The article highlights computing, talent, energy, and industrial capacity as the foundations of AI power. Control over semiconductors, data centers, and advanced manufacturing shapes who can develop and deploy leading systems. At the same time, public policy plays a critical role through export controls, research funding, infrastructure investment, and regulatory signals that influence private-sector behavior.
Moreover, the authors stress that risk management is not a constraint but a source of legitimacy and stability. Managing safety, misuse, and escalation risks is essential to sustaining public trust and avoiding destabilizing outcomes as AI capabilities grow.
Diffusion, Allies, and Global Competition
Diffusion emerges as a central theme. The systems that spread first will shape global norms, economic gains, and governance standards. While China actively promotes its AI systems abroad, the authors argue that the United States must better integrate AI diffusion into its foreign policy by working with allies and offering trusted alternatives.
In many scenarios, competition is less about who invents first and more about who deploys faster and at scale. Partnerships with allies are therefore critical to ensuring that democratic values influence the global AI ecosystem.
Strategy Across Multiple Futures
Finally, the article calls for a probabilistic approach to AI strategy. Policymakers should identify a base-case scenario but stress-test policies against more dangerous alternatives. The goal is not perfect prediction but resilience—building strategies that remain effective as conditions change.
The authors conclude that AI should be treated not as a single story, but as a shifting landscape. Strategic success will depend on the ability to adapt, hedge risks, and align innovation with national and global interests.
Reference
Sullivan, J., & Feldman, T. (2026). Geopolitics in the age of artificial intelligence: Strategy and power in an uncertain AI future. Foreign Affairs. https://www.foreignaffairs.com