Environmental Sustainability in the Era of Artificial Intelligence: Energy Use, E-Waste and Emerging Challenges

Artificial intelligence (AI) is rapidly becoming a general-purpose technology, but its environmental footprint is also expanding through electricity demand, data-centre construction, semiconductor production, cooling, water use and accelerated turnover of electronic equipment. This paper evaluates environmental sustainability in the era of AI through a life-cycle framework focusing on energy consumption, electronic waste (e-waste), water demand, critical minerals, rebound effects and emerging governance challenges. The study applies a descriptive and analytical secondary-data methodology using evidence from the International Energy Agency (IEA), International Telecommunication Union (ITU), United Nations Institute for Training and Research (UNITAR), United Nations Conference on Trade and Development (UNCTAD), United Nations Environment Programme (UNEP), Organisation for Economic Co-operation and Development (OECD), and peer-reviewed studies published through 2025. The IEA estimated global data-centre electricity consumption at about 415 TWh in 2024, roughly 1.5% of global electricity use, and projected approximately 945 TWh by 2030 in its base case, with AI as the most important contributor to growth. Global e-waste reached 62 million tonnes in 2022, while only 22.3% was documented as formally collected and environmentally soundly recycled; the total is projected to reach 82 million tonnes by 2030. The analysis shows that efficiency gains alone cannot guarantee sustainability because growing demand, increasingly complex models, short hardware replacement cycles and rebound effects may offset resource savings. A sustainable AI pathway therefore requires transparent life-cycle measurement, low-carbon electricity, efficient and task-appropriate models, water-sensitive infrastructure, extended hardware life, circular-economy systems and environmental standards that distribute benefits and burdens more equitably.