Abstract
In this essay, I explore the potential implications of adolescent use of AI assistants on the development of adolescents’ capacities for autonomy. Thinking through these issues from the perspective of first principles is, I believe, needed in order to sharpen our understanding of some of the issues AI assistants pose for young people. A clear analysis will suggest which legal or policy prescriptions should follow.
I first give an overview of adolescent use of AI assistants, noting their popularity and the variety of use contexts. Drawing upon the work of Catriona Mackenzie, I then outline a relational conception of autonomy that conceives of autonomy as having three distinct dimensions: self-determination, self-governance, and self-authorization. The precision of Mackenzie’s conception allows us to more carefully consider how different uses and contexts of technological use can affect individuals’ capacities for autonomy. Using this three-dimensional conception of autonomy, I propose three potential interventions for law and policy: thinking beyond harms, substantive privacy protections, and autonomy impact assessments.