The authors synthesize observations from 26 Infinite Potential games modeling AI economic disruption dynamics. They identify the issues, capabilities and playbooks, and recommendations that participants believe necessary to navigate a period when social legitimacy, economic stability, and technical control might collapse. The findings aim to help policymakers, researchers, and the public understand the preparations needed before such an event.
Topline
AGI-driven economic and social disruptions could outpace every tool the government currently has to respond.
Artificial general intelligence (AGI) remains hypothetical, but its potential to disrupt economies, displace workers, and concentrate power is already forcing hard questions that policymakers aren’t prepared to answer.
RAND’s Infinite Potential platform uses scenario-based tabletop games to explore how governments and institutions might respond to artificial intelligence– and AGI-driven crises. This report synthesizes findings from 26 Economic Crisis and Business and Civil Society scenario games conducted between April 2025 and May 2026 with participants from government, business, and civil society.
Key Takeaways
Labor displacement could overwhelm existing safety nets.
Players across nearly every game warned that AGI could displace entire categories of workers far faster than retraining programs, unemployment systems, or job programs can respond.
The U.S. tax system isn’t built for an AI economy.
When productivity growth decouples from employment, income tax revenue shrinks while demand for public spending surges; this is a fiscal dynamic that existing tax structures aren’t designed to handle.
Frontier AI firms might accumulate power rivaling that of governments.
Players repeatedly raised concerns that a small number of companies controlling transformative AI could resist oversight, capture political processes, and concentrate wealth in ways that destabilize democratic governance.
Competition between the United States and China creates unresolvable tension.
Accelerating AI adoption to keep pace with China could worsen domestic instability, but slowing down risks losing economic and geopolitical ground. Players found no easy answer.
Recommendations
Policymakers should anticipate AGI threats before crises force their hand.
- Develop real-time labor-market monitoring and pre-authorized emergency fiscal tools that can activate when displacement thresholds are crossed.
- Build government capacity to independently evaluate frontier AI systems rather than relying on AI lab self-reporting.
- Create playbooks for managing relationships with dominant AI firms to guide issues from information-sharing to emergency intervention.
- Develop and test new approaches to how people earn, contribute, and are supported under AGI, such as pilot programs for reduced work weeks, universal basic income, national service frameworks, and care-economy expansion, before a crisis makes deliberation impossible.
- Pursue international coordination on AI taxation and trade measures to prevent capital flight and regulatory arbitrage.
Methodology
RAND researchers designed and ran 26 scenario-based tabletop games between April 2025 and May 2026 using RAND’s “Day After” methodology, which confronts players with a plausible crisis and asks them to reason through its implications and responses. Games typically included between ten and 16 players and lasted two hours. Players included former government officials, business and civil-society leaders, AI policy experts, and think tank researchers from the United States, Europe, and Asia. A standardized notetaking template captured player discussions; findings were synthesized using a secure, RAND-internal AI tool and then reviewed for accuracy and completeness.
Topics
This publication is part of the RAND research report series. Research reports present research findings and objective analysis that address the challenges facing the public and private sectors. All RAND research reports undergo rigorous peer review to ensure high standards for research quality and objectivity.
This document and trademark(s) contained herein are protected by law. This representation of RAND intellectual property is provided for noncommercial use only. Unauthorized posting of this publication online is prohibited; linking directly to this product page is encouraged. Permission is required from RAND to reproduce, or reuse in another form, any of its research documents for commercial purposes. For information on reprint and reuse permissions, please visit www.rand.org/pubs/permissions.
RAND is a nonprofit institution that helps improve policy and decisionmaking through research and analysis. RAND's publications do not necessarily reflect the opinions of its research clients and sponsors.