UK AI Data Centres Are Expanding: What Could It Mean for Energy and Water?

Artificial intelligence is moving further into film, television, visual effects, animation and digital creation, but the tools appearing on screen depend on a much larger physical system behind them. As demand for AI grows, the UK is encouraging the development of new data centres capable of providing the computing power needed by businesses, researchers and creative industries.

That expansion could bring investment, jobs and greater access to powerful technology, including for smaller entertainment companies and independent creators. It also raises practical questions about how much electricity and water future data centres could need, where they should be built and whether existing infrastructure can keep pace.

Why is the UK building more AI data centres?

AI systems require significant computing power, particularly when major models are being trained or demanding applications such as image and video generation are used at scale. Increasing the amount of computing available in Britain could allow UK businesses and creative industries to access powerful AI systems without depending entirely on facilities overseas.

By January 2026, the UK Government said five AI Growth Zones had been designated across Great Britain. These areas are intended to support new AI infrastructure by bringing together access to land, electricity, planning and investment, with the Government linking the programme to £28.2 billion of investment and more than 15,000 jobs.

For entertainment, the connection is already developing. British Film Institute research published in June 2025 found generative AI being explored in areas including visual effects, dubbing, animation and production workflows, while independent creators are increasingly using AI to produce images and video directly for audiences.

How much electricity could data centres need?

The scale of interest in new data centre development is considerable. Great Britain’s energy regulator, the Office of Gas and Electricity Markets, known as Ofgem, said in July 2026 that applications for large electricity connections had risen from 41 gigawatts to 125 gigawatts in less than a year, with data centre projects accounting for at least 80 gigawatts of that total.

A gigawatt, shortened to GW, measures power at a particular moment rather than electricity used over an entire year. For comparison, Government discussions earlier in 2026 referred to Great Britain’s peak electricity demand as being around 45 gigawatts.

That does not mean future data centres are about to consume almost twice the country’s peak electricity demand. Connection applications represent interest rather than guaranteed future demand, and Ofgem has warned that many projects in the queue may be speculative or never built.

The figures instead show just how much potential development is being proposed. Ofgem is now looking at tougher requirements intended to remove projects that are unlikely to proceed and free up grid capacity for developments that are genuinely ready to connect.

Could AI put pressure on the electricity grid?

Potentially, although the scale of the impact will depend on how many projects are actually built, where they are located and how quickly the electricity network expands. Large data centres need reliable connections capable of delivering significant amounts of power, which can create challenges in places where the grid is already heavily used.

The Government’s AI Energy Council has been examining whether some computing work could become more flexible. This could include shifting certain tasks to different times or locations when electricity demand is lower, although not every AI workload can simply be paused whenever the grid becomes busy.

There is also an important carbon question. The Government’s April 2026 modelling estimated that UK AI computing could be responsible for between 34 and 123 million tonnes of carbon dioxide emissions between 2025 and 2035, with the unusually wide range reflecting uncertainty over both the growth of AI and how quickly Britain’s electricity system becomes lower carbon.

What about water?

Electricity is only part of the picture because powerful computer processors produce heat and need cooling. Some data centres can use substantial amounts of water, although demand varies greatly according to the size of the facility, the cooling technology being used and the climate in which it operates.

England’s Environment Agency says some large data centres can consume millions of litres of water each day. Closed loop cooling systems can reuse water and substantially reduce ongoing demand, while evaporative cooling systems can consume more water because some is lost during the cooling process.

This variation is important because there is no single amount of water used by every data centre. A facility designed around water efficient cooling could place very different demands on local supplies from another centre of a similar size using a different system.

The wider planning system is also still adapting. Data centre demand is not yet fully factored into national water demand planning, meaning future growth could introduce additional requirements that were not included when some longer term water resource plans were developed.

Where data centres are built could matter

The location of future AI infrastructure may become almost as important as the technology inside it. Areas with suitable electricity connections, reliable water supplies and greater access to lower carbon energy may be better equipped to support very large data centres.

Planning decisions could therefore involve more than simply finding enough land. Local electricity capacity, water availability, cooling requirements and the environmental impact of building and operating a facility could all become increasingly important considerations.

The Government is trying to address some of these questions through AI Growth Zones and its wider compute strategy. The aim is to expand Britain’s AI capacity while also considering how new facilities connect to electricity networks and other essential infrastructure.

Why should the entertainment industry care?

A television production company or independent creator using an AI video tool may never know which data centre processed their request. The environmental cost of the computing does not disappear simply because the servers are located somewhere else.

For example, generating several AI versions of the same visual effects shot can increase the number of times a model has to process a request, which in turn increases the computing required. If that type of workflow becomes common across film, television, advertising and digital production, millions of individual requests could collectively create significant demand.

There are potential benefits too. AI could make some creative processes faster, reduce certain production costs and give smaller teams access to tools that previously required much larger budgets, while more efficient processors could reduce the electricity required for individual tasks.

The British Film Institute has argued for an approach that allows the screen sector to explore those benefits while developing clearer information around sustainability. That balance is likely to become more important as generative AI moves from experimentation into everyday production workflows.

Can Britain expand AI more sustainably?

There is no reason that every new data centre has to create the same environmental impact. More efficient processors, improved cooling, better use of lower carbon electricity and careful decisions about location can all make a difference.

The challenge is that AI demand is growing quickly at the same time as the UK is trying to expand electricity networks, improve water resilience and reduce carbon emissions. Building more AI infrastructure therefore involves balancing technological and economic ambitions with the physical limits of the systems that have to support it.

For entertainment, this may seem far removed from the finished film, television programme or AI generated video. Yet as creative industries rely more heavily on artificial intelligence, the electricity grids, cooling systems and data centres behind those tools will become an increasingly important part of understanding what AI powered entertainment really requires.

For more on how artificial intelligence is changing film, television, music and digital storytelling, explore SoapLandTV’s AI & Entertainment page.

Sources: Ofgem data from July 2026, UK Government AI and compute research from April 2026, Environment Agency water planning material and British Film Institute and CoSTAR research from June 2025.