AI is positioned as a transformative asset for sustainability, yet recent research highlights serious hidden costs. A systematic review in Sustainability analyzes 138 studies from 2018-2024, linking AI to the UN Sustainable Development Goals (SDGs) while cautioning against an “AI green paradox,” where ecological harms may counterbalance technological benefits. AI enhances sustainability in sectors like energy, agriculture, and environmental monitoring; enabling smart grids, precision farming, and waste sorting.
However, significant risks exist, including high energy consumption for AI operations, algorithmic opacity, and socio-economic disparities that could deepen the digital divide. This “digital pollution” incorporates issues of access and trust, hampering AI’s potential for positive impact.
To ensure AI aids sustainability, recommendations include promoting “Green AI,” enhancing data governance, and developing educational programs for responsible AI deployment. Regulatory frameworks are also essential for enforcing ethical standards, ensuring that AI serves social and environmental priorities rather than exacerbating existing challenges.
