University of Manchester experts give evidence on environmental impact of AI and data centres

Susan
2 days ago
2 min read

Researchers from The University of Manchester are advising parliament on the growing energy and environmental impacts of artificial intelligence (AI) and data centres, as part of a new inquiry into their implications for the UK’s net zero ambitions.

Data centres have been designated as critical national infrastructure due to their importance for economic growth, but their electricity consumption is projected to quadruple by 2030. The inquiry will assess how this increasing demand could affect energy and water systems and how emerging technologies and policy approaches could reduce environmental impacts.

In their evidence, Dr Alejandro Gallego Schmid and Dr Raphael Tarpani, researchers at the University’s Tyndall Centre for Climate Change Research, highlight a number of challenges associated with this growth, including:

Rising carbon emissions from both electricity use and the manufacturing of hardware.
Increasing demand for critical materials such as copper, silicon and rare elements.
Growing volumes of electronic waste driven by rapid hardware replacement cycles.
Potential strain on water resources and local environments.

They argue that current policies do not yet fully account for the pace and scale of AI driven demand and recommend:

Integrating data centre growth into wider energy, infrastructure and environmental planning, ensuring expansion is aligned with grid capacity and the availability of low carbon electricity.
Improving transparency around environmental impacts through better reporting of energy, water and material use, alongside accounting for full lifecycle of digital infrastructure, such as hardware production, supply chains and electronic waste.
Supporting a circular economy approach to digital technologies, promoting the reuse, repair, refurbishment and recycling of servers and other hardware to reduce resource demand and waste.
Managing the resource pressures associated with AI and data centre expansion, including demand for critical minerals.

The evidence highlights emerging technologies that could reduce environmental impacts, including more efficient chips, advanced cooling systems and ‘green AI’ approaches that limit unnecessary computation.

The researchers also point to opportunities for data centres to contribute to local energy systems, for example, by recovering waste heat to supply homes and buildings, or by providing flexibility to help balance electricity demand.

Dr Alejandro Gallego Schmid said, ‘Data centres are fundamental to the digital economy and will play an important role in enabling AI innovation. However, their expansion needs to be planned alongside the UK’s wider sustainability objectives.

‘Our evidence shows that solutions are available but many of these will require investment in infrastructure and more coordinated action across policy, industry and research.’

The submission has been supported by Policy@Manchester, the University’s policy engagement unit.

Read the full written submission here: Written evidence - DCU0023