Algorithmic Challenges to Autonomous Choice

Abstract

We then explore how autonomous algorithmic assistants affect the legal framework. Some issues challenge the very use of algorithmic assistants: Should the law place an age limit on the use of such algorithms? Should legal limits be placed on their use in certain spheres? Other issues arise from the construction of preferences by algorithms. To wit, in a market in which demand is driven by algorithms, can we necessarily regard the choices made as expressions of user preferences that serve our socio-political goals? Even if the answer is positive—does this new mode of user choice fit current legal assumptions, such as those that apply to the notions of consent or intent? And are our regulatory tools, which seek to ensure that individual consumers can make informed decisions, outmoded? It is therefore essential to determine whether the existing legal framework is sufficiently potent to deal with this brave new world, or whether we need new regulatory tools. Despite their importance and timeliness, these questions have not been dealt with in depth. This article seeks to fill this void. It proceeds as follows. Part II explores the lure and modes of operation of algorithmic assistants, and how these characteristics may affect human choice. Part III then analyzes the rationales for such choice and explores how these rationales are affected by the employment of autonomous algorithmic assistants. As shown, while some rationales are not harmed—and might even be strengthened— by the use of autonomous algorithmic assistants, others challenge us to reconsider the meaning and the role that choice plays in our lives and to deal with the conflict between the efficient fulfillment of short-term preferences and the long-term ability to form such preferences. Part IV analyzes the implications of these new private orderings on regulation, with a special focus on laws based on assumptions of human autonomous decision-making. We summarize our findings in a short conclusion.