TITLE:
Artificial Intelligence and Consumer Decision-Making: AI-Powered Personalization, Trust Pathways, and Purchase Intentions in Nigeria
AUTHORS:
Nwachukwu Ugochukwu Goodness Oluoma, You Yi
KEYWORDS:
Artificial Intelligence, Personalization, Cognitive Trust, Affective Trust, Purchase Intention, Moderated Mediation, Privacy Risk, Nigeria, Emerging Markets
JOURNAL NAME:
Open Access Library Journal,
Vol.13 No.8,
August
20,
2026
ABSTRACT: Artificial intelligence (AI)-powered personalization is reshaping how consumers in emerging digital economies encounter, evaluate, and commit to market offerings, yet prevailing scholarship continues to treat personalization as a uniformly trust-enhancing stimulus. This study challenges that assumption by advancing a capability-trust contingency model in which the effect of AI-powered personalization on purchase intention is neither direct nor uniform but routed through two analytically distinct trust pathways and conditioned by consumer beliefs about machine competence and data exposure. Drawing on initial trust theory and the personalization-privacy paradox, the model specifies cognitive and affective trust in the AI agent as parallel mediators, with perceived AI capability moderating the personalization-to-cognitive-trust path and privacy risk perception moderating the personalization-to-affective-trust path. The framework is examined in Nigeria, Africa’s largest consumer market and a context where rapid mobile-commerce diffusion coexists with weak institutional data-protection enforcement, conditions under which Western trust assumptions cannot be presumed to transfer. Using survey data from 412 active online shoppers and a moderated-mediation estimation strategy, the analysis demonstrates that personalization elevates purchase intention primarily by building cognitive trust, that this indirect effect strengthens markedly as perceived AI capability rises, and that privacy risk perception attenuates the affective-trust pathway. The contribution is threefold: it decomposes a previously monolithic trust mechanism, it identifies the boundary conditions that govern when personalization helps or backfires, and it extends consumer AI theory to an under-studied institutional setting. Implications for adaptive personalization strategy, transparency signalling, and trust calibration in low-institutional-trust markets are discussed.