Abstract: The rapid advancement of generative artificial intelligence (Generative AI) is reshaping consumer behavior, yet its systemic impact on consumer cognition and behavior remains under-explored. This research builds a "technology - cognition - behavior" interdisciplinary framework to investigate the dual-edged effects, mechanisms, and strategies of Generative AI usage on consumer behavior, focusing on the dynamic interplay between short-term memory efficacy enhancement and long-term cognitive hindrance. The research unfolds through four progressive studies: Study 1 explores the nonlinear impact of Generative AI usage on consumer memory efficacy; Studies 2 and 3 analyze the positive and neutral behavioral effects driven by memory efficacy enhancement and the negative consequences caused by memory efficacy impairments, respectively; Study 4 identifies effective marketing intervention strategies to counteract the negative effects revealed in Study 3. Theoretical contributions include bridging the gap in dual-edged effect research from a dynamic perspective, extending the applicability of metamemory theory to technological contexts, and providing a novel paradigm for interdisciplinary theoretical integration. Practical implications involve offering feasible solutions for balancing efficiency and ethical risks in AI product development, supplying scientific evidence for consumer cognitive protection and policy formulation, and fostering ethical technology development aligned with societal well-being.