AI-Powered Social Media Platforms and Their Effect on Emotional Well-Being

Abstract

The rise of social media applications based on artificial intelligence created by the use of AI has completely changed the cyber-communication environment as it has raised a lot of questions concerning the psychological effects of using these applications among the younger population groups, including college students. This paper has explored how the use of AI-driven social media platforms can influence the emotional health of students at a university in Lahore, Pakistan. A cross-sectional survey design was used as a quantitative research design whereby 420 students who were selected through convenience sampling across six universities (three public and three private) in Lahore were used to collect the data. The warwick-edinburgh mental well-being scale (WEMWBS) was applied to measure emotional well-being, and self-reported frequency and engagement indicators on platforms such as Tik Tok, Instagram Reels, YouTube Shorts and Snapchat were used to measure AI-powered platform use. The instruments were adjusted to the Pakistani cultural situation, and all of the items were found to be very internally consistent (.78-91 Cronbach alpha). Data analysis was done using demographic analysis, descriptive statistics, Pearson correlation and partitioning through structural equation modeling (SEM) using AMOS 26 and SmartPLS 4.0. The findings showed that passive content consumption with AI involvement and emotional well-being were significantly negatively correlated (beta = -.34, p <.001), whereas active and creative intervention showed positive correlations (beta =.22, p <. 01). Several pathways were modulated with age, gender and type of university. The structural model had good fit indices (CFI =.96, RMSEA =.054, SRMR =.062). Policymakers in digital literacy, mental health and responsible AI design policy in higher education in Pakistan can use these findings to implement changes in these areas.