Privacy-Aware Personalization: Machine Learning, Model Validation, And Consumer Trust in AI-Mediated Marketing

Abstract

This study has examined how privacy-aware personalization, machine learning personalization effectiveness, AI model validation and reliability, and privacy The increasing use of artificial intelligence (AI) and machine learning in personalized marketing has enhanced organizations’ ability to deliver relevant and individualized consumer experiences while simultaneously intensifying concerns regarding privacy, transparency, algorithmic reliability, and consumer control. This study examines the effects of Privacy-Aware Personalization Capability (PAPC), Machine Learning Personalization Effectiveness (MLPE), AI Model Validation and Reliability (AIMVR), and Privacy Transparency and Consumer Control (PTCC) on Consumer Trust in AI-Mediated Marketing (CTAIM). Grounded in Privacy Calculus Theory, the study employed a quantitative, explanatory, cross-sectional survey design. A total of 330 questionnaires were distributed, yielding 300 usable responses from consumers with experience of AI-supported personalized marketing. Data were analyzed using descriptive statistics, reliability and validity assessment, Pearson correlation analysis, and multiple linear regression. The measurement instrument demonstrated strong internal consistency, with Cronbach’s alpha coefficients ranging from .86 to .91, while sampling adequacy was confirmed by a KMO value of .902 and a significant Bartlett’s Test of Sphericity (p < .001). The regression model explained 72.4% of the variance in consumer trust (R² = .724; adjusted R² = .720; F(4, 295) = 193.46, p < .001). AI Model Validation and Reliability emerged as the strongest predictor of consumer trust (β = .33), followed by Machine Learning Personalization Effectiveness (β = .27), Privacy Transparency and Consumer Control (β = .24), and Privacy-Aware Personalization Capability (β = .21), with all relationships statistically significant at p < .001. The findings support all four hypotheses and demonstrate that consumer trust in AI-mediated marketing depends not only on personalization effectiveness but also on dependable algorithmic performance, responsible data practices, transparency, and meaningful consumer control. The study extends Privacy Calculus Theory by highlighting perceived algorithmic reliability as an important uncertainty-reducing mechanism in consumers’ evaluation of AI-driven personalization.