Are AI Recommendations Less Green? The Effects of AI Recommendation Sources and Green Product Attribute Framing on Consumer Responses

ABSTRACT Extant research has reported mixed evidence regarding the effectiveness of artificial intelligence (AI) recommendations in promoting green products, leaving it unclear under which conditions such recommendations are more (or less) effective. Drawing on Self‐Determination Theory (SDT), this research addresses this ambiguity by proposing a congruence effect between AI recommendation sources (substitute versus augmented AI) and green product attribute framing (utility‐based versus sustainability‐based). Using four scenario‐based online experiments (total N = 1114)—including one product choice task, one preference evaluation, and two studies measuring purchase intention and willingness to pay—this research demonstrates that substitute (augmented) AI recommendations are more effective for utility‐framed (sustainability‐framed) green products. Furthermore, consumers' sense of control mediates this congruence effect, and providing attribute‐based post hoc explanations eliminates the disadvantage of substitute AI under sustainability framing by shifting consumer responses toward sustainability‐framed green products. These findings advance understanding of human–AI interaction in green consumption and offer actionable insights for businesses seeking to leverage AI recommenders to increase the market share of green products.