Consumer Privacy Concerns and Acceptance of AI-Driven Targeted Advertising

AI-driven targeted advertising infers consumers’ interests, circumstances and intentions from their data and serves advertising accordingly, and its acceptance depends on how consumers weigh its benefits against the privacy it consumes. This paper models acceptance through a privacy calculus in which privacy concern reduces acceptance, perceived benefit raises it, and perceived control and transparency both raise acceptance directly and, in the case of control, weaken the effect of concern. A simulation of 1,000 consumers shows that privacy concern is the largest determinant (−0.42), followed by perceived benefit (+0.31), perceived control (+0.27), transparency (+0.21) and perceived creepiness (−0.19); the concern–control interaction is positive (+0.20), so that the slope of acceptance on concern is −0.36 under low control and −0.25 under high. Among highly concerned consumers, acceptance is 0.39 with low control and 0.65 with high, close to the 0.71 of consumers with low concern. The paper argues that control is the lever that reconciles targeting with privacy concern, that transparency and benefit matter but do not substitute for it, and that creepiness, the response to inferences consumers did not expect, is a distinct cost that targeting design should avoid. Implications for advertisers, platforms and regulation are derived.