Environmental Perception Shaped By Recommendation Algorithms: A Study on Information Relevance and Human-AI Interaction

This study explores the impact of digital recommendation algorithms on environmental awareness, focusing on the dynamic interplay among perceived information relevance (PIR), algorithmic content exposure (ACE), and human–machine interaction (HMI). Utilizing the Elaboration Likelihood Model and Uses and Gratifications Theory, the study reconceptualizes awareness as a dynamic product of cognitive, technological, and behavioral dimensions in AI-mediated contexts. A quantitative survey (N = 385) stratified by platform behavior across Southeast Asia revealed that PIR has the strongest predictive value (β = 0.55), followed by ACE (β = 0.49). HMI functions as a key moderator (β = 0.57), suggesting that interactive engagement significantly strengthens cognitive assimilation of environmental content. These results suggest that AI platforms, when ethically designed, can serve as vehicles for ecological education. The findings call on platform engineers, sustainability advocates, and policymakers to implement value-sensitive design strategies, prioritize relevance-based content, and embed interactive features. Practical implications emphasize the role of value-sensitive design and interactive engagement in environmental communication. Limitations include reliance on self-report data and regional scope. Recognizing limitations in sampling and self-reported data, future research should employ longitudinal and behavioral methodologies to capture long-term attitudinal shifts. Overall, this study contributes to the integration of environmental psychology with algorithmic ethics by offering a robust, empirically grounded framework for digital ecological cognition.