The rapid diffusion of generative artificial intelligence (AI) has transformed consumer interaction with digital systems, shifting AI use from an exploratory novelty to a routine component of everyday tasks. While existing research has primarily focused on initial adoption and perceived usefulness of AI technologies, far less is known about how consumer motivation and interaction behavior evolve with repeated use. This study investigates how consumers’ motivational states and effort investment change over the course of repeated interactions with generative AI. Drawing on consumer motivation and effort allocation perspectives, the research examines whether continued exposure leads to sustained engagement or systematic behavioral adjustment. Using interaction-level data that track consumers’ sequential engagements with generative AI systems, the study distinguishes between early-stage and later-stage interactions within usage sessions, and examines changes in behavioral indicators of effort, including prompt length and the time interval between successive interactions. The results show that later-stage interactions are characterized by significantly shorter prompts and shorter inter-interaction intervals, indicating reduced cognitive effort and faster interaction pacing rather than disengagement. These patterns suggest that consumers adapt their interaction strategies toward more streamlined and economical use as experience accumulates. Importantly, the findings exhibit pronounced temporal heterogeneity. Interaction routinization intensifies as generative AI usage becomes more mature and normalized. This study extends consumer research on human–AI interaction by moving beyond adoption-centric models and conceptualizing motivation as an experience-dependent and temporally evolving construct. From a managerial perspective, the findings highlight the importance of adaptive AI interface design that supports efficient and routinized interaction, rather than relying solely on novelty or early-stage performance improvements to sustain long-term use