ALGORITHMIC INSECURITY: HOW DIGITAL PERSONALIZATION THREATENS EMOTIONAL STABILITY

Abstract

This study examines how algorithm-driven personalized recommendations influence emotional recovery, focusing on the relationship between algorithmic exposure and users’ emotional stability. The research explores the construct of algorithmic insecurity, which describes how personalization systems on social media and music platforms may reinforce negative affective states rather than support emotional regulation. A cross-sectional design was employed with 440 participants (M = 22.60 years, SD = 4.73), including 219 males (49.8%), 214 females (48.6%), 6 non-binary (1.4%), and 1 undisclosed (0.2%). Participants completed self-report measures assessing their emotional states and the nature of personalized content exposure. Findings revealed that music algorithms predominantly suggested sad or melancholic content during distressing periods, significantly more than expected (χ²[2, N = 440] = 294.99, p < .001). Longer exposure to personalized content was significantly associated with higher frequencies of sadness or distress (χ²[12, N = 440] = 69.70, p < .001), while social media exposure to mood-congruent content correlated with greater emotional instability (χ²[16, N = 440] = 64.61, p < .001).These results demonstrate that algorithmic personalization can perpetuate low mood and emotional dysregulation by continuously presenting mood-congruent content, effectively sustaining users’ negative affect. The study highlights algorithmic insecurity as a psychological risk inherent in personalization systems and underscores the need to integrate emotional security into discussions of digital well-being and online safety.