Autonomy, Paternalism, and AI-Driven Nudges

Nudging is a morally contentious method of influence. One concern is that nudges threaten to undermine individual autonomy. With the rise of artificial intelligence systems, this worry has only magnified, with some arguing that we should be even more concerned about the threat that AI-driven nudges pose to our autonomy than the threat that traditional, “human-driven” nudges pose. Call this the comparative threat claim (CTC). Calboli (Philosophy& Technology 38, 2025) argues that to evaluate the plausibility of the CTC, we should distinguish between two dimensions of autonomy that AI-driven nudges threaten to undermine: self-agency and autocracy. I have two aims in this discussion. The first is to build on Calboli’s conceptual framework by introducing a third dimension of autonomy that is relevant to concerns about autonomy and nudging: sovereignty. The second is to evaluate the plausibility of the CTC with respect to paternalistic nudges, in particular. I argue that, for this kind of nudge, the CTC is implausible when it comes to a nudgee’s autocracy and sovereignty. Overall, the discussion contributes to our understanding of how concerned we should be about the impact that AI-driven nudges have on our autonomy.