Author(s) :

Emily Han.

Volume/Issue :

Abstract :

As AI capabilities have accelerated over the past decade, so have the questions surrounding their environmental impact. This narrative review examines how perceptions of AI’s environmental impact have evolved across academia, industry, and public discourse between 2014 and 2025. The focus is on three key eras of development. Drawing on peer-reviewed literature, corporate sustainability reports, and public publishing outlets, a qualitative Sentiment Concern Index (SCI) framework was used to interpret shifts in optimism and concern across academia, industry, and publishing houses. The findings suggest that while early academic and industrial discourse framed AI as a promising but untested tool, more recent years have seen both increased deployment and growing criticism, especially regarding the energy demands of large-scale models. Despite these concerns, the landscape is shifting toward “green AI,” carbon-aware infrastructure, and environmentally responsible development practices. The paper concludes with a forward-looking discussion of integrated strategies, emphasizing the need for coordinated policy, technical innovation, public transparency, and cross-sector collaboration. As AI becomes further embedded in society, ensuring that it functions as a climate asset and not an environmental liability will be one of the defining sustainability challenges of the coming decade.