Adverse Environmental and Public Health Effects of Artificial Intelligence: A Narrative Review

The rapid global expansion of artificial intelligence (AI), particularly generative models, drives energy-intensive data centers with substantial environmental and public health costs. This narrative review synthesizes information obtained from the scientific literature confirming contributions of AI to greenhouse gas emissions, freshwater depletion, e-waste, and air pollution from fossil-powered grids. Public health risks include algorithmic bias exacerbating disparities, AI-generated misinformation/deepfakes eroding trust, privacy loss, mental health harms, and job displacement impacting social determinants. These burdens disproportionately affect marginalized communities via environmental justice failures and biased algorithms. While acknowledging that certain AI applications, particularly in climate modeling, medical diagnostics, and energy optimization, may offer net benefits under appropriate governance, this review focuses on the documented adverse impacts of current large-scale, commercial AI deployment patterns._ _Mitigation demands life-cycle assessments, renewable energy mandates, circular hardware economies, bias audits, and policies prioritizing health equity. Sustainable AI requires coordinated action across stakeholders. Implementing these mitigation strategies is constrained by major obstacles in cost, technical infrastructure, and governance. Overcoming these barriers requires the development of comprehensive economic analyses and structured strategic roadmaps.