ABSTRACT Artificial intelligence has become consumers' primary decision‐making resource, raising two questions: how do consumers justify accepting AI as a trusted source of reasoning, and when does this acceptance maintain rather than forfeit their capacity to think independently? Drawing on Bandura's moral disengagement theory, we introduce AI epistemic disengagement , a self‐regulatory process through which consumers neutralize the discomfort of delegating reasoning to AI and justify reliance on it. In two studies with nearly 600 US consumers, we find two distinct patterns of AI dependence: augmentation and substitution. Augmentation maintains epistemic responsibility through conscientious evaluation, preserving consumer agency. Substitution abdicates epistemic responsibility through uncritical acceptance, making consumers vulnerable. These patterns differ in their antecedents and outcomes, consistent with Self‐Determination Theory's distinction between autonomous and controlled motivation. Augmentation is associated with Need for Cognition and prompting skills, while the reverse holds for substitution. Augmentation, rather than substitution, is associated with intense AI use, habit formation, and broader application. Developing prompting skills, promoting epistemic responsibility, and designing interfaces that encourage critical thinking can help consumers, platforms, and policymakers preserve consumer welfare while supporting AI adoption.
Theoharakis et al. (Tue,) studied this question.