Artificial Intelligence and Its Growing Environmental Impact: Challenges and Sustainable Solutions

Artificial intelligence is being adopted across nearly every sector, yet the computing infrastructure behind it draws growing amounts of electricity and water and depends on resource intensive hardware that eventually becomes electronic waste. At the same time, artificial intelligence tools help forecast weather, balance power grids, and monitor ecosystems. This paper reviews the evidence on both sides of that balance. Using a descriptive and analytical secondary research design, it draws on peer reviewed studies, intergovernmental reports, and official statistics from India, the United States, and international agencies, published mostly between 2019 and 2026. The reviewed sources show that data centre electricity demand is projected to more than double by 2030, that water use occurs both on site and through power generation, that hardware manufacturing contributes embodied emissions that are often left out of reporting, and that electronic waste is growing much faster than documented recycling. The sources also indicate that the environmental benefits of artificial intelligence applications are real but depend on deployment context and are seldom measured with the same rigor as the costs. The paper classifies impacts across the life cycle, compares costs with benefits, and proposes a layered Green AI framework covering governance, infrastructure, models, and applications. Recommendations are given for developers, operators, researchers, and policymakers. No new experiments were conducted, and the findings are limited by the quality and comparability of published data.