Author Name: Akhil Date: 27-03-2026
The deployment of artificial intelligence in media recommendation systems has enabled the construction of granular psychological profiles of users based on behavioral trace data, content interaction patterns, and passive digital signals. This paper provides a comprehensive theoretical and empirical analysis of how AI-driven recommendation systems infer, represent, and operationalize psychological constructs — including personality traits, emotional states, cognitive styles, and motivational orientations — to optimize content delivery and behavioral prediction. Drawing on the Five-Factor Model of personality, the OCEAN-social media inference literature, and algorithmic profiling theory, the paper examines the psychometric validity, ethical implications, and democratic consequences of AI-mediated psychological profiling at scale. The Cambridge Analytica case is analyzed as a paradigmatic instance of psychographic targeting that collapsed the boundary between psychological research and political manipulation. The paper reviews the empirical evidence for personality inference from digital footprints including Facebook likes (N = 58,000; r = .56 for openness), social media language predictions, and behavioral clickstream data, and critically evaluates the gap between prediction accuracy and psychological construct validity. A taxonomy of profiling mechanisms is proposed distinguishing explicit psychometric input, behavioral inference, network-structural inference, and temporal pattern extraction. The paper argues that AI psychological profiling constitutes a form of epistemic asymmetry where platforms know more about users psychological states than users themselves, with profound implications for autonomy, consent, and cognitive liberty. Regulatory frameworks including the EU AI Act, GDPR Article 22, and India’s Digital Personal Data Protection Act are evaluated for their adequacy in addressing algorithmic psychological profiling.
Keywords: AI profiling; recommendation systems; personality inference; OCEAN model; psychographic targeting; algorithmic profiling; digital footprints; cognitive liberty.
