Hungary is facing historic water shortages and an energy crisis, which increases the importance of using resources efficiently. Recent Microsoft case studies illustrate that artificial intelligence (AI) can both reduce energy and drinking-water consumption and create new operating demands. The company emphasizes that advanced AI systems can generate savings, while their own operation — especially data centres and large models — requires measures to limit added water and energy use.
Concrete examples and figures
- Using an AI-based leak-detection solution, Microsoft says it saved more than 8.5 million cubic metres of drinking water last year — a volume comparable to the annual residential water use of a city the size of Debrecen.
- At Microsoft’s Redmond (Puget Sound) campus, an AI-driven fault-detection system achieved roughly 4,000 megawatt-hours (MWh) of energy savings in 2025.
- Since 2024 the company has circulated cooling water in closed systems directly around chips; this approach reduces evaporation-related losses and has saved more than 125 million litres of water per data centre per year.
Reducing data-centre water use
Microsoft is deploying next-generation, low-water cooling methods in its Azure data centres. In 2025 it tested approaches that direct coolant precisely to the chip areas that need it (so-called microchannel cooling), which can improve cooling efficiency up to threefold in some setups. The company tailors water-supply solutions to each site’s local climate and natural conditions; for example, in Amsterdam they have largely used precipitation water for server cooling.
Savings in building operations and workload management
- In building management, IoT and predictive AI systems monitor external conditions in real time to regulate energy and water use; Microsoft cites LinkedIn campuses and offices as examples of optimized facility operations.
- For several Copilot products, notably GitHub Copilot, Microsoft uses dynamic model selection to match compute resources to task complexity, reducing unnecessary consumption.
Relevance to Hungary’s water losses
Hungary’s supply networks experience substantial water losses: on average about one quarter of piped drinking water does not reach consumers, and in Budapest the loss is approximately 16%, largely due to leakage. This underlines the potential relevance of AI-based leak detection for domestic water management.
Water return and corporate targets
- Microsoft reports a 25% improvement in its water-use effectiveness (WUE) since 2022 and has set a target of 40% improvement by 2030.
- In one year the company returned 14.2 million cubic metres of water to the environment, and for the first time the volume returned exceeded the volume withdrawn that year.
- The total contracted volume of water the company has committed to return reaches 133 million cubic metres, roughly equivalent to the annual residential water consumption of about 18 cities the size of Debrecen.
Partnerships and watershed modelling
Working with the International Water Management Institute (IWMI), Microsoft developed Water Copilot — a digital twin of a river basin that provides comprehensive information on water stocks, supporting irrigation planning and preparation for water-scarce periods.
Dual nature of the technology and open questions
Microsoft argues that AI applications can cut operational costs and emissions over both the short and long term, a point echoed by Bábel Gabriella, CEO of Microsoft Hungary, who noted that advanced technology enables sustained energy and water savings. At the same time, it remains necessary to weigh AI’s own energy and water footprint against these efficiency gains. The showcased measures — leak detection, optimized facility control, closed-loop and microchannel cooling — indicate progress in reducing water use, but the net effect across the full AI infrastructure will depend on further technical developments and location-specific solutions.
Related event
The Portfolio Sustainable World 2026 conference will take place on 8 September; the sector-neutral event will address green-economy issues, including water retention in Hungary.
(More details are available in Microsoft’s press materials and sustainability report.)
