AI’s Use of Knowledge in Society

Hayek’s famous insight was that central planning—even if economically efficient—is not feasible because the necessary knowledge is inherently dispersed throughout the economy. The advent of transformative AI demands a reappraisal of Hayek’s “knowledge problem” and its implications for how decision rights are allocated within firms and society. We develop a property rights framework in which powerful AI shifts the optimal locus of control through two channels: (i) by codifying local knowledge that was previously tacit and inalienable, and (ii) by expanding the information processing capacity of agents to aggregate, interpret, and act on data. These forces erode the informational advantage of maintaining on-the-spot decision-makers, making centralized coordination and control more feasible and more efficient—especially where complementarities across assets are important. The framework yields several predictions: larger average firm size, greater industry concentration, and reduced local managerial autonomy. We review early evidence and find that it is largely consistent with these patterns. We also discuss conditions that can still favor decentralization. The implications of our analysis extend beyond economic considerations: centralization of economic power can lead to centralization of political power and dampen incentives to invest in human capital.