Automated Governance

Abstract

Although one might contend that acquiescing to algorithmic decisions as a consumer is a personal and private choice, the question remains whether, as citizens, we should also relinquish control to algorithmic decision-making for state governance and oversight over laws. Our trepidation over governmental algorithmic decision-making relates to our visceral dread of the inflexibility of machines—an obduracy that can wreak fatal consequences. However, although we may continue to fear algorithmic overlords, the ship has already sailed on the debate over automated governance. A path-breaking report chaired by Professors David Engstrom, Daniel Ho, Catherine Sharkey, and Justice Mariano-Florentino Cuellar provides a comprehensive overview of the automated decision-making that governmental agencies have already deployed and show that nearly half of the federal agencies studied (forty-five percent) have delved into artificial intelligence (“AI”) and related machine learning (“ML”) tools. First, the Article notes the benefits of automated governance, which include: (1) efficiency, (2) cost-savings, and (3) the capability to uncover historical patterns of bias. Thus, the only true remaining questions are: (1) Whether there are still meaningful differences in human versus algorithmic decision-making such that some scenarios should preclude the use of the latter; and (2) whether there is a right to explanation and contestation regarding automated governance? In addition to answering these questions, the Article also offers proposals for how the Equal Employment Opportunity Commission and Federal Trade Commission may make new use of automated tools to prevent racial and other types of discrimination in business decision-making. Finally, the Article details several guardrails that should be put in place to ensure that automated governance will serve the greater good.

Keywords: AI, Artificial Intelligence, Administrative Law, Automated Decision-Making, AI Hiring, AI and Employment, EEOC, FTC, Federal Trade Commission, Equal Employment Opportunity Commission, AI and Bias, Work Law, Workplace, Discrimination, Antidiscrimination