An AI-Driven Intelligent System for Early Detection of Gaming Addiction and Harmful Digital Behaviour Among Indian Adolescents

Abstract India’s adolescent population forms one of the largest digital gaming communities worldwide. While gaming provides cognitive stimulation, prolonged and uncontrolled participation can evolve into behavioural addiction, accompanied by toxic online interactions. This research develops and evaluates an AI-driven framework for early detection of gaming addiction and harmful digital behaviour among Indian adolescents aged 13–19 years. A mixed-method design combined behavioural surveys with artificial-intelligence-based textual and temporal analytics. Using data from 1,200 students across Delhi NCR, Maharashtra and Karnataka, a hybrid CNN–BERT model was trained on chat sentiment, screen-time patterns and aggression indicators. Quantitative analysis using Pearson correlation revealed a significant association between gaming duration and aggression (r = 0.71, p < 0.01). The system achieved 94 percent accuracy with an F1 score of 0.91 for risk classification. Results highlight AI’s potential as an early-warning instrument for digital-wellness promotion. The study recommends embedding such predictive frameworks within India’s NEP 2020 policy ecosystem to strengthen mental-health and digital-safety initiatives. Keywords: Artificial Intelligence, Gaming Addiction, Toxic Behaviour, Adolescents, Machine Learning, CNN-BERT, Digital Well-Being, India, Behavioural Analytics, Education Policy