A new UNU-INWEH report warns that AI-related infrastructure could consume 9.3 trillion litres of water annually by 2030, equivalent to the domestic water needs of approximately 1.3 billion people for a year. The study highlights that AI’s environmental footprint extends beyond carbon emissions to include significant impacts on water resources, land use, and growing electricity demand from data centres. For businesses, the findings reinforce the need to consider AI’s environmental impacts within broader ESG, climate, and digital transformation strategies.
A new United Nations study has raised concerns about the growing environmental footprint of artificial intelligence (AI), warning that the rapid expansion of AI infrastructure could consume enough water each year to meet the domestic needs of approximately 1.3 billion people by 2030. The findings highlight a critical but often overlooked sustainability challenge as businesses, governments, and investors accelerate AI adoption across the global economy.
Published by the United Nations University Institute for Water, Environment and Health (UNU-INWEH), the report, The Environmental Cost of AI’s Energy Use: Carbon, Water and Land Footprints, estimates that AI-related infrastructure could consume approximately 9.3 trillion litres of water annually by the end of the decade. While AI is widely recognised for its potential to improve productivity, automate processes, optimise resource use, and support sustainability initiatives, the report emphasises that the technology itself carries a rapidly growing environmental cost.
AI’s Hidden Water Footprint
Water consumption is emerging as one of the most significant environmental impacts associated with AI. The majority of this demand originates from data centres, which require large volumes of water to cool servers and high-performance computing systems that power AI models and applications.
According to the report, AI’s projected annual water footprint of 9.3 trillion litres would be sufficient to cover the domestic water requirements of approximately 1.3 billion people for an entire year. As AI adoption continues to accelerate, water availability may become an increasingly important factor in data-centre location decisions, infrastructure planning, and corporate sustainability strategies.
The findings are particularly relevant for regions already experiencing water stress, where growing demand from digital infrastructure could place additional pressure on local freshwater resources.
Beyond Carbon Emissions: The Full Environmental Cost of AI
The report argues that AI should no longer be evaluated solely through its energy consumption and carbon emissions. Instead, organisations should assess AI through a broader environmental lens that includes water use, land requirements, and the wider impacts of supporting infrastructure.
Researchers estimate that carbon emissions linked to AI-related energy consumption could approach 400 million tonnes of CO₂ annually by 2030. At the same time, expanding data-centre infrastructure will require increasing amounts of land for facilities, electricity transmission networks, and associated infrastructure.
As a result, the environmental footprint of AI extends across multiple sustainability dimensions, creating challenges that many organisations have yet to incorporate into their climate and ESG strategies.
Data Centres Become a Growing Sustainability Challenge
The study projects that electricity and water consumption from data centres could roughly double by 2030 as AI workloads continue to increase. AI is expected to account for an increasingly large share of global data-centre resource demand, driven by the rapid growth of generative AI, machine learning, cloud computing, and advanced analytics applications.
Environmental impacts vary significantly depending on location. Data centres powered by fossil-fuel-intensive electricity grids generate substantially higher carbon emissions, while facilities located in water-scarce regions can exacerbate local water shortages.
This means that the sustainability performance of AI depends not only on technological efficiency improvements but also on energy sourcing, cooling technologies, infrastructure design, and geographical context.
What This Means for Businesses
For businesses, the report highlights a new dimension of AI governance and sustainability risk. AI is increasingly embedded across operations, supply chains, software development, customer services, ESG reporting, and decision-making processes. However, the environmental impacts associated with AI infrastructure may become an increasingly important consideration within climate strategies, Scope 3 emissions assessments, supplier due diligence, and broader sustainability reporting frameworks.
Investors and regulators are also paying growing attention to the environmental impacts of digital infrastructure, particularly as demand for AI continues to accelerate globally. Organisations may face increasing expectations to disclose and manage the resource consumption associated with their digital operations and AI deployments.
The report therefore calls for greater transparency and standardised reporting approaches covering AI-related carbon emissions, water consumption, and land use. As companies continue to invest heavily in AI-driven transformation, understanding these environmental trade-offs will become increasingly important for responsible business strategy and long-term sustainability performance
Read more at: