Ethical Implications of AI Personalization in Everyday Decision-Making

Thesis Abstract

Digital platforms increasingly tailor content and recommendations to individual users, shaping everyday choices in areas ranging from consumer behavior to political information. This study investigates the ethical implications of AI personalization in everyday decision-making, addressing concerns about autonomy, informed consent, transparency, power asymmetries, and potential bias. The aim is to assess how personalized AI systems influence decision processes and to explicate the normative boundaries that ought to govern such systems. Specific objectives are (1) to examine users’ perceptions of autonomy and control in the context of personalized recommendations; (2) to analyze how transparency and explainability affect trust and acceptance of AI-driven suggestions; (3) to identify discriminatory or unfair outcomes arising from algorithmic personalization across demographic groups; (4) to evaluate the relationship between data practices (collection, usage, and consent) and users’ moral agency; and (5) to develop a normative framework informed by ethics of care, virtue ethics, and Kantian autonomy to guide responsible design and policy. The study adopts a mixed-methods design, combining a cross-cultural survey with in-depth interviews to triangulate quantitative and qualitative insights. The population comprises adult internet users in the United States, United Kingdom, and Nigeria, with a target sample of 2,400 survey respondents selected via stratified random sampling to ensure representation across age, gender, ethnicity, and socioeconomic status. In-depth interviews will be conducted with 36 participants drawn from survey respondents who indicate substantial reliance on AI personalization in daily decisions, ensuring diversity in digital literacy and platform use. Data collection instruments include a structured questionnaire measuring perceived autonomy, perceived transparency, trust in AI, and perceived fairness, alongside a semi-structured interview guide exploring experiences of personalization, data practices, and perceived moral obligations of platforms. Validity and reliability will be established through pilot testing (n=120), Cronbach’s alpha analysis for multi-item scales (target ? ? 0.70), and test-retest reliability. Documented platform policies and terms of service will be analyzed to triangulate stated consent with actual data practices. Statistical analysis will employ multiple regression to test hypotheses about the effects of perceived autonomy and transparency on trust in AI and decision satisfaction, controlling for demographic variables. Moderation analyses will examine whether digital literacy and platform type moderate these relationships. Mediation analysis will assess whether perceived autonomy mediates the relationship between personalization intensity and perceived fairness. Thematic analysis will be applied to interview transcripts to extract nuanced moral and contextual factors influencing ethical evaluations, using a six-phase framework (familiarization, coding, theme development, reviewing, defining, and reporting). The study will integrate normative theory with empirical findings by mapping results onto an ethics of care perspective (focusing on relational autonomy and vulnerability), virtue ethics (character formation and deliberative flourishing), and Kantian autonomy (universal moral law and rational agency), to derive a composite normative framework for governance and design. Expected findings include (a) nuanced evidence that high personalization can both enhance convenience and erode perceived autonomy, with transparency and user control emerging as critical moderators; (b) differential impacts across cultural contexts and demographic groups, suggesting potential bias in personalization pipelines; (c) a robust link between inadequate data consent practices and diminished user moral agency; and (d) convergent themes indicating a need for more transparent explanation of why particular recommendations are made, accompanied by user-centered controls over data-sharing and algorithmic settings. The study contributes to knowledge by integrating empirical social science methods with normative philosophy to illuminate how AI personalization affects moral agency and social equality, offering a disciplined framework for evaluating and regulating personalization practices. Practical recommendations will include design guidelines for explicable personalization, consent-centered data practices, and policy interventions that promote fairness, accountability, and relational autonomy. The main conclusion posits that ethically responsible AI personalization requires not only technical transparency but granular control over data practices and a normative commitment to preserving individuals’ substantive autonomy and social dignity. Recommendations target platform designers, policymakers, and researchers, advocating standardized transparency disclosures, opt-in personalization schemas, ongoing impact assessments, and interdisciplinary collaboration to monitor ethical outcomes across diverse user populations.