Algorithmic Nudging and Financial Over-Indebtedness: A Longitudinal Analysis of AI-Integrated BNPL Schemes in E-Commerce: Evidence from MENA

The rapid proliferation of artificial intelligence-integrated “buy now, pay later” (BNPL) schemes in Middle Eastern and North African e-commerce has intensified concerns regarding consumer financial vulnerability. This study examines how algorithmic nudging, AI personalization intensity, and perceived ease of credit influence financial outcomes through impulsive buying tendency, with financial literacy evaluated as a moderating protective factor. Employing a multi-method design, we applied partial least squares structural equation modeling to cross-sectional data from 1,247 BNPL users across seven MENA countries, supplemented by a six-month longitudinal follow-up tracking debt accumulation and financial stress. Results demonstrate that algorithmic nudging strongly predicts impulsive buying, which subsequently serves as the primary driver of financial stress and debt. While AI personalization enhances platform loyalty, it concurrently exacerbates financial risk through indirect pathways. Crucially, financial literacy significantly buffers the adverse effects of both algorithmic nudging and impulsive buying, reducing longitudinal debt trajectories by 47% among highly literate consumers. These findings advance theoretical frameworks on digital choice architecture and consumer vulnerability, supporting evidence-based regulatory interventions including nudge-transparency mandates, contractual cooling-off periods, and systematic credit-bureau reporting.