Technology Governance in the Age of AI: From Redistribution to Predistribution

Abstract

What should we do about labor automation in the age of artificial intelligence? Early proposals such as Universal Basic Income focused on taxing and redistributing the proceeds of AI. Now, an emerging slate of ideas tagged to the concept of “predistribution” proposes distributing ownership through citizen stakes in AI or sovereign wealth funds. But what is “predistribution”? This Article presents a positive and unifying theory of predistribution, describes its categories, and explains how it can remake work in the age of AI. For over a century, Americans have focused on tax-and-transfer redistribution to address income inequality in the realm of consumption while calibrating tax policy to produce the right mix of forward incentives for production. Predistribution encompasses the ways of intervening in the realm of production before market outcomes are produced for consumption. This Article defines “predistribution” as the distribution of power over economic outcomes and identifies four categories of predistribution: (1) incentives and bargaining power, (2) public production, (3) ownership, and (4) governance and control. For the question of how to govern technology and labor automation, the Article turns to the fourth category: governance. Because governing technology at work begins with governing work itself, it examines the law of labor and technology. Surveying the twentieth-century history of U.S. labor and employment law, the Article finds that labor law under the National Labor Relations Act (NLRA) places decisions about technology beyond the scope of worker governance while employment and tax law incentivize labor automation. To address this picture, it advances a model for the future of work based on predistributive governance—distributing control to democratize decisions about technology in the workplace.