Marketing safety, producing risk: a Peltzman effect framework for AI in consumer health

Artificial Intelligence-enabled health technologies are positioned as instruments of risk reduction, yet this framing obscures a counterintuitive possibility: the safety signals marketers attach to these products may encourage riskier behaviour. We draw on the Peltzman Effect, risk compensation theory’s central insight that safety interventions can provoke offsetting behaviour, to develop a conceptual framework for how AI-mediated safety cues reshape consumer health behaviour. We propose a three-stage pathway in which the gap between claimed and objective safety lowers perceived risk, expands behavioural latitude, and produces unintended harm. The framework distinguishes AI modalities, differentiates preventive and compliance behaviours, specifies boundary conditions, and recasts algorithmic reassurance as moral hazard. Testable propositions reposition AI-enabled safety as a behavioural signal rather than a straightforward product benefit.