The Algorithmic Well‐Being Paradox: When AI Decision Delegation Helps and Hurts Consumer Well‐Being

Abstract

ABSTRACT As generative AI increasingly shapes consumer decision‐making, its implications for consumer well‐being remain unclear. Existing research largely focuses on the trade‐off between algorithmic efficiency and autonomy loss. This research advances understanding by proposing the algorithmic well‐being paradox, whereby the same AI delegation can enhance or diminish well‐being depending on the consumption context. Integrating Savoring Theory and Self‐Determination Theory, we develop a framework that explains when and why these divergent outcomes occur. To test the proposed relationships, four studies were conducted with 2 × 2 between‐subjects experiments ( N = 1427) spanning multiple consumption domains, including cameras, guided tours, ergonomic chairs, glamping trips, smartphones, Arctic expeditions, and smart‐home systems. Study 1 demonstrates that the effects of algorithmic delegation differ between experiential and material consumption. Studies 2 and 3 identify anticipatory savoring and autonomy frustration as the underlying mechanisms, while Study 4 confirms the proposed dual‐mediation process in a high‐complexity context. Results show that consumption type shapes how consumers respond to AI delegation. In experiential consumption, AI delegation increases well‐being by enhancing anticipatory savoring, with these benefits outweighing autonomy concerns. In material consumption, AI delegation reduces well‐being by increasing autonomy frustration, which diminishes the benefits of convenience. These findings advance human‐AI interaction research by demonstrating that the impact of algorithmic delegation depends on consumers' underlying consumption goals and provide guidance for designing more human‐centered AI decision‐support systems.