Higher education is transitioning from mass expansion to high-quality development. In this process, mental health issues among college students have become increasingly prominent, encompassing not only academic stress-induced anxiety but also complex challenges such as interpersonal adaptation difficulties and career planning confusion. Traditional “problem-oriented” intervention models have shown limitations in responsiveness and adaptability, often only passively addressing existing psychological crises rather than preventing them in advance. This study aims to explore an AI-powered “positive psychology” proactive intervention model through developing an intelligent system. The system automatically collects, filters, and personalizes recommendations for positive activities on campus. Using a randomized controlled trial design, we conducted an 8-week intervention study involving 126 college students at a university. The study found that AI-based “campus positive activity” recommendations effectively boost students’ positive emotions and promote psychological capital development through cumulative micro-interventions. This provides universities with empirical evidence and innovative methods to implement low-cost, efficient, and scalable mental health promotion programs through smart technology.