The Coasean Singularity? Demand, Supply, and Market Design with AI Agents (www.nber.org)

A new NBER working paper, "The Coasean Singularity? Demand, Supply, and Market Design with AI Agents," maps how autonomous AI agents—systems that search, negotiate, and transact for humans—could reshape digital markets by sharply reducing transaction costs. The authors frame agent uptake as derived demand: principals trade off improved decision quality against lower effort, with adoption driven by agent capability and the task’s context. On the supply side, firms face strategic choices about designing, integrating, and monetizing agents, and whether agents operate inside single platforms or across ecosystems—choices that change competition, data flows, and pricing power. Technically, agents lower search, communication, contracting, preference elicitation, identity verification, and enforcement costs, expanding the set of feasible market designs and enabling new transaction types. But they also introduce frictions (congestion, coordination costs, and price obfuscation) and novel regulatory issues around transparency, liability, and platform governance. The paper stresses the net welfare effects are empirical, not theoretical certainties, and calls for focused economic research to guide policy and market design as AI-mediated transactions scale.