Agency in the Age of AI

TL;DR

This paper addresses the societal harms of generative AI and proposes agency as the central analytic lens. It outlines a blue-sky program to extend the Planning Theory of Agency and the Belief-Desire-Intention model, integrate relational and enactive perspectives, and quantify agency changes using information-theoretic methods. The approach envisions agent-based models that mix baseline humans, AI-augmented humans, and autonomous AI agents to simulate social scenarios like elections or epidemics, enabling scenario-specific interventions. The work aims to provide a principled framework for evaluating harms and guiding the design of AI tools and policies to augment rather than undermine human agency.

Abstract

There is significant concern about the impact of generative AI on society. Modern AI tools are capable of generating ever more realistic text, images, and videos, and functional code, from minimal prompts. Accompanying this rise in ability and usability, there is increasing alarm about the misuses to which these tools can be put, and the intentional and unintentional harms to individuals and society that may result. In this paper, we argue that \emph{agency} is the appropriate lens to study these harms and benefits, but that doing so will require advancement in the theory of agency, and advancement in how this theory is applied in (agent-based) models.