Answers are grounded in this paper and cite their sources.
Deepfakes and generative AI are reshaping advertising, yet research remains fragmented across marketing, communication, psychology, and information systems. Through an integrative review of 67 articles, we develop the concept of synthetic persuasion, the process through which AI-generated or AI-manipulated media shapes consumer responses. We advance a four-stage advertising influence pipeline spanning production and representation, signaling and inference, evaluation and behavior, and diffusion and societal consequences. Synthetic origin tightens dependencies between stages that conventional theory treats as separable, so production choices shape inference and individual responses aggregate into institutional consequences. Research concentrates in the middle stages, while production systems and societal consequences remain underdeveloped. We extend theory by adding disconfirmation as a mechanism within the Persuasion Knowledge Model, anchoring authenticity to multiple cues instead of a human author, and accounting for the decoupling of perceptual quality from production effort. Seven propositions follow, each addressing an unresolved tension in the pipeline.
No takes yet. Share an insight, caveat, or question.
Sands et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: