Consumer Trust in AI-Generated Marketing Content: A Systematic Literature Review and Research Agenda

Purpose. As generative artificial intelligence transforms marketing content production, questions about consumer trust have become urgent for both scholars and practitioners. This study systematically reviews the empirical and conceptual literature on consumer trust-related responses to AI-generated marketing content (AIGC), mapping antecedents, mediating mechanisms, moderating conditions, and downstream outcomes. Design/methodology/approach. Following PRISMA 2020 reporting principles, a systematic search of Scopus and Google Scholar identified 59 records. After deduplication, screening, and eligibility assessment, 35 studies published between 2020 and 2026 were retained. Each study was coded along eleven dimensions including methodology, theoretical framework, independent and dependent variables, mediators, and moderators. Thematic synthesis organized findings into four themes: the disclosure dilemma, authenticity as central mediator, moderating conditions, and downstream outcomes. Findings. AI disclosure activates persuasion knowledge and erodes trust-related outcomes across diverse marketing contexts, yet these effects are neither universal nor uniform. Perceived authenticity emerges as the primary mediating mechanism, with moral disgust operating as a parallel affective pathway. Content type (emotional versus rational), consumer AI literacy, cultural context, anthropomorphism cues, and disclosure framing moderate the strength and direction of effects. An integrative input-process-output framework synthesizes these findings. Originality/value. To the authors' knowledge, this review provides one of the first focused syntheses of empirical and conceptual work on the trust construct in AI-generated marketing content. It proposes an integrative framework organizing antecedents, mediators, moderators, and outcomes, and advances five research propositions addressing habituation dynamics, cross-cultural variation, content-modality contingencies, disclosure framing, and behavioral measurement gaps.