RAND Advises U.S. to Keep Options Open on Superintelligence Path

RAND published a paper recommending that the United States preserve flexibility as it approaches superintelligence, since the transition's shape remains uncertain. The proposed approach rests on four areas: human-AI collaboration and safety, security and verification architecture, modernizing national security institutions, and helping citizens, companies, and governments adapt. It also lays out seven possible strategic archetypes in three families, including dominance, co-development, and preparedness.

RAND's paper argues that uncertainty about how a superintelligence transition unfolds makes a fixed strategy risky, so it recommends preserving freedom of action. Its four pillars are a human-AI ecosystem, an AI-security architecture, modernized national security institutions, and greater response capacity for citizens, firms, and governments.

It also maps seven strategic archetypes across three families: coexistence, denial, and acceleration. Coexistence includes dominance, co-development, and preparedness; denial includes moratorium, deterrence, and continuity of society; acceleration relies on markets, competition, and rapid iteration. The paper further flags uncertainties such as danger proximity and whether stable human-AI coexistence is feasible.