The Privacy-Utility Trade-Off in Artificial Intelligence Contexts - A Systematic Literature Review

The adoption of Artificial Intelligence (AI)-driven technologies offers significant benefits for digital ecosystems, including Digital Identity Management (IdM), by enabling personalization and enhanced decision-making. However, this simultaneously amplifies privacy risks through extensive data collection and inference. This Systematic Literature Review (SLR) investigates how research conceptualizes the trade-off between privacy and utility in AI contexts. We applied an AI-augmented review methodology based on Wolfswinkel et al.’s guidelines to analyze 133 peer-reviewed articles. Our synthesis reveals that the trade-off is conceptualized across five interconnected themes: (1) theoretical foundations challenging rational choice models; (2) limitations of privacy-enhancing technologies (PETs); (3) domain-specific implications in healthcare and IoT; (4) human dimensions of trust and transparency; and (5) regulatory governance. We identify a critical disconnect between technical privacy research and behavioral/organizational realities. The paper concludes with recommendations for integrative theoretical models and standardized evaluation metrics to better balance privacy protection and data utility in AI-driven identity ecosystems.