Summary
- ING identifies data centres, renewable energy, and related infrastructure as a major engine of sustainable financing.
- Green bonds and loans are expected to account for about 60% of sustainable-debt issuance this year.
- Greater scrutiny of data-centre community impacts could make sustainable financing more selective.
Data-centre construction is becoming a larger part of the sustainable-finance market as the capital required for AI infrastructure converges with grid investment, renewable generation, and tighter environmental scrutiny.
ING Wholesale Banking said data centres, renewable energy, and related infrastructure have emerged as an engine of sustainable financing and expects that momentum to continue through 2026 and 2027.
The bank’s latest global sustainable-finance outlook argues that growing electricity demand is creating financing requirements across the infrastructure chain rather than at data-centre sites alone. New capacity can require grid reinforcement, generation, substations, storage, and other energy assets in addition to the buildings and IT equipment themselves.
ING expects green bonds and green loans to remain the dominant products in the sustainable-debt market, together accounting for around 60% of global issuance this year. That compares with around 50% in 2024 and 40% in 2021.
Sustainability-linked financing has moved in the opposite direction. ING said sustainability-linked bond and loan issuance fell 54% year on year in the first half of 2026 to just over $60bn, as borrowers with eligible investment programmes increasingly favoured green use-of-proceeds structures.
Data centres present an unusually complicated test for those instruments. Facilities can support digitalisation and AI workloads while adding large, continuous electrical demand to local power systems. The environmental case therefore depends heavily on what is financed, how energy is sourced, how efficiently the site operates, and whether the infrastructure creates additional pressure on constrained grids or water systems.
ING expects scrutiny of community impacts to make sustainable debt for data centres more selective. Those impacts can include electricity-system reinforcement, noise, land use, water consumption, backup generation, and the distribution of economic benefits from large developments.
European regulation is also moving towards more detailed measurement of facility performance. Energy and water reporting requirements are increasing the amount of operational data available to policymakers and investors, while future efficiency standards could make technical performance more relevant to financing terms and eligibility.
That raises the bar for developers seeking to attach a green label to financing. Renewable-energy contracts can address part of the emissions profile, but lenders and investors may increasingly examine cooling design, PUE, WUE, embodied carbon, backup systems, heat reuse, and whether grid investment is incorporated into the overall development strategy.
The financing need is substantial because AI infrastructure has pushed project sizes upwards. Large campuses can require several billion pounds or euros across land, shells, mechanical and electrical plant, grid connections, and customer fit-out. The result is growing dependence on project finance, corporate debt, infrastructure funds, private credit, and securitisation alongside conventional developer equity.
Sustainable finance can lower the distance between those capital pools and projects that can demonstrate measurable environmental performance, but the category is unlikely to provide a shortcut around weak infrastructure fundamentals. A site without a credible power path or with poorly understood water and community impacts remains exposed regardless of the financing label attached to it.
As more data-centre investment enters the market, the distinction between financing a digital asset and financing an energy-intensive industrial facility is becoming harder to sustain. The debt structure increasingly has to account for the same power, construction, environmental, and operating risks that determine whether the campus can be built and run in the first place.
