A Systems-Level Analysis of Algorithmic Regulation

Abstract

Algorithmic tools are being wielded in the name of regulatory values as regulatory tools are lagging in mitigating the impacts of algorithmic systems. In this thesis, I characterize and evaluate the systemic relationship between regulation and algorithmic technologies in two parts. In Part I, I uncover the current mismatched application of laws to algorithmic systems and propose resulting implications and mitigations. In Part II, I consider regulatory design for emerging technologies that incentivizes efforts toward increasing the foreseeability of harm.

While each chapter centers the interplay between different regulations and algorithmic technologies, the problems that are uncovered and the solutions proposed generalize to reasoning about algorithmic regulation as a whole. This analysis highlights the unexpected ways that regulations can shape incentives for algorithmic development, as well as the unexpected ways that algorithmic innovation can spark regulatory innovation.