Algorithmic allure: a theoretical framework for dark AI patterns and the erosion of informed consent in AI-driven digital marketing

Abstract

This study examines how generative AI is reshaping digital marketing by enabling unprecedented levels of personalization, automation, and consumer interaction, while simultaneously producing subtler, more complex forms of algorithmic manipulation that extend well beyond conventional dark patterns and threaten meaningful informed consent. Prior research has treated AI-enabled persuasion in a fragmented way—focusing on personalization, recommender systems, or deceptive interface designs in isolation—without offering an integrated account of how generative AI shapes influence across the consumer journey. To address this gap, we conduct a multi-case qualitative analysis of seven documented international incidents of AI-driven marketing practices from 2024 to 2026. Data sources include regulatory reports, investigative journalism, corporate disclosures, and peer-reviewed literature. Using reflexive thematic analysis, we identified seven high-level themes: identity manipulation, social-proof exploitation, synthetic evidence, AI-washing, agentic AI manipulation, synthetic authenticity, and autonomous commercial manipulation. These themes form a taxonomy of “Dark AI Patterns”. Building on these findings, we propose Algorithmic Allure Theory (AAT), which conceptualizes AI-enabled persuasion as a dynamic, six-stage process: observation, interpretation, algorithmic allure, behavioral steering, delegation, and algorithmic governance. AAT argues that contemporary commercial persuasion relies on adaptive computational systems that continuously generate, test, and refine influence tactics in real time—processes that often erode informed consent while preserving a veneer of legitimacy and transparency. Contributions (1) establish Dark AI Patterns as a distinct class of deceptive practices; (2) introduce AAT as a transdisciplinary framework for understanding AI-mediated influence; (3) reconceptualize informed consent as an ongoing cognitive and temporal process rather than a one-time legal event; (4) offer a practical taxonomy to guide future research, regulation, and ethical AI marketing practices As a primary goal.