The technology sector is experiencing rapid growth, largely accelerated by the development of artificial intelligence (AI). This expansion is driving an unprecedented demand for data centre infrastructure and semiconductor manufacturing capacity. Global spending on AI data centres is estimated to approach USD 7 trillion by 2030, with total investment in AI infrastructure projected to reach USD 31.6 trillion by 2050. This growth is intensifying the sector’s pressure on freshwater systems, as both data centre cooling and semiconductor fabrication are highly water-intensive activities, often located in geographies already facing acute resource stress.

Introduction

The technology value chain spans a sequence of activities including critical mineral mining, ultra-pure water-intensive semiconductor fabrication, hardware assembly, and the operation of data centres. Each stage carries distinct nature-related dependencies and impacts. For example, growing demand for advanced chips increases land use change, water withdrawal, and wastewater discharge. Climate change compounds these pressures, creating an inextricable link where interventions to reduce water impacts, such as switching to air cooling, can inadvertently raise energy use and emissions.

Financial institutions: integrating nature in decision making

Financial institutions like Norges Bank Investment Management (NBIM) and ‘Company A’ are increasingly evaluating their portfolios’ exposure to water-related issues. For NBIM, technology is its largest sector, representing 33% of its listed equity portfolio. These organisations use a mix of in-house tools and third-party datasets, such as the WWF Water Risk Filter and WRI Aqueduct, to assess risks. Analysis found that indirect water consumption, primarily for electricity generation, can represent up to 75% of a data centre’s total consumption. NBIM’s analysis revealed that while less than 6% of known assets are located near ecologically sensitive locations, sector-level aggregation can mask significant company-level variation.

Corporates: using the LEAP approach to assess, manage and disclose nature-related issues

NTT DATA and Vanguard International Semiconductor Corporation (VIS) applied the TNFD LEAP (Locate, Evaluate, Assess, Prepare) approach to identify nature-related issues. NTT DATA screened 141 global data centres, identifying 23 facilities with potentially high water risk. It validated these results using ThinkHazard! to account for potential overestimations in urban basins. VIS, which operates several fabrication plants (fabs), identified that while baseline water stress at its fab locations is generally low, drought remains a priority physical risk. VIS updated its environmental risk management policy in 2024 to explicitly include nature-related risks alongside climate-related risks.

In the Evaluate phase, companies estimated water-related metrics to support their assessments. NTT DATA expanded its water monitoring coverage from 78.3% in FY2023 to 87.7% in FY2024, using Water Usage Effectiveness (WUE) as a key indicator. VIS has set targets of greater than 70% for its fabs’ water process recovery rates, with three out of five facilities targeting 85%. These recovery rates are achieved through low-consumption process machines and process water recovery systems. The marginal cost of achieving higher rates increases as more expensive measures are required once cost-effective options are exhausted.

Challenges and potential solutions when undertaking nature-related assessments

Participants encountered significant challenges regarding data availability and granularity. NBIM reported that asset-level data is only partially available, and many company assessments focus only on direct operations rather than the full value chain. NTT DATA noted that its co-location model creates structural data access problems, as clients may own equipment and procure energy directly, limiting access to energy and water data. Furthermore, a lack of industry consensus on metrics like WUE complicates peer benchmarking and the setting of meaningful targets.

Conclusion

Water-related issues are consistently identified as material across the technology sector and are recognised as an industry-wide challenge. Addressing these issues is increasingly seen as a strategic driver of resilience and a potential business opportunity. The interdependence between financial institutions and corporates is critical; institutions rely on companies for comparable data to support capital allocation decisions. The report concludes that integrating nature and climate assessments is vital to avoid unintended consequences when managing trade-offs between water consumption and energy demand.