Tech esg

The UK AI and Net-Zero equation

Jason Mitchell · Posted on: February 20th 2026 · read

In September 2021, the UK published its National AI Strategy, a 10-year plan aimed at strengthening the UK as a global centre for AI innovation and leadership.

More recently, in a letter to the UK energy secretary, Labour MP Toby Perkins, who chairs the Environmental Audit Committee, said data centres are a ‘key area of concern’ in hitting the UK’s net-zero emissions target.

Data centres are essential to the operation of digital services such as streaming, and artificial intelligence (AI).

This potentially creates an enormous, and largely unreported, dilemma for a government celebrating AI as a major platform for future economic growth and the creation of jobs.

The UK government has supported plans for more data centres to be built, to help turn the country into an ‘AI superpower’.

Conversely, according to Toby Perkins MP, the UK government is relying on an emissions-reduction plan, the Seventh Carbon Budget, that makes no allowance for the impact of data centres.

According to the BBC, almost 100 new data centres were being planned in 2025.

The issue highlights a trend that appears to be repeatedly dogging climate change initiatives: the often-experienced trade-off between addressing environmental harm and driving for economic growth.

Rollback of many climate -related policies in the United States, in order to promote US industries, is another example.

Conversely, according to Toby Perkins MP, the UK government is relying on an emissions-reduction plan, the Seventh Carbon Budget, that makes no allowance for the impact of data centres.

The UK government’s Compute Roadmap forecasts a three-fold increase in AI-capable data centre capacity by 2030 to support advanced AI workloads. Broader industry analysis suggests data centre electricity demand might grow significantly, largely driven by AI demands, unless sustainable measures are put in place. Without decarbonised energy, the demand could push carbon emissions upwards, risking the UK’s net-zero by 2050 commitment.

 

A further issue is that measuring the true impact of AI usage is not straightforward: many UK firms lack accurate emissions data for AI and data centre operations at present.

How then might the AI and net-zero equation be balanced? The answer lies in taking decisive action.

The UK is aiming to build more clean energy capacity to support higher electricity demand while reducing emissions. This includes major projects such as the expansion of nuclear generation as part of its energy security and net-zero strategy.

"The UK government has also established an AI Energy Council to coordinate the energy infrastructure needed for AI growth alongside net-zero ambitions. The Council is working on grid upgrades, connections, and sustainable power options for data centres."

Jason Mitchell, Partner

"Initiatives such as AI Growth Zones aim to couple planning certainty with energy and economic benefits, accelerating access to power without compromising environmental rules."

Mark Lumsdon Taylor, Executive Development & Sustainability Lead

In parallel, data centre operators are increasingly adopting advanced cooling and energy-efficient designs to reduce wasted power and environmental impact.

Some UK startups are working to anchor AI infrastructure directly to clean energy sources, potentially decoupling computation expansion from fossil fuels completely.

Making environmental reporting mandatory for data centres could help. This would enable energy use, water use, and carbon emissions to be tracked and managed.

Pricing and subsidies could also prioritise clean energy sources for data centres, encouraging renewable procurement and storage that matches computing demand spikes.

AI itself could be used to optimise grid performance, including balancing supply and demand, forecasting renewable generation, and reducing waste to help the grid cope with higher overall loads.AI itself could be used to optimise grid performance, including balancing supply and demand, forecasting renewable generation, and reducing waste to help the grid cope with higher overall loads.

Many analysts also argue that nuclear capacity, and other reliable low-carbon baseload sources will be essential if industrial and AI-related electricity demand grows sharply.

Indeed, there has been considerable research and development relating to small modular reactors (SMRs), attached to data centres as a means to power energy-hungry processes such as AI.

On the surface, it would appear that things are well under control to balance AI growth and net-zero achievement in the UK.

In response to the question from Toby Perkins MP, the Department for Energy Security and Net Zero (DESNZ), explained to the BBC that emissions from data centres ‘are factored into our modelling’.

2050

A new carbon budget will be published in June 2026, as part of the government’s legally binding commitment to achieve net-zero emissions by 2050. The Budget is expected to use forecasts of electricity demand that reflect wider economic trends.

However, according to Mary Stevens, innovation lead and AI campaigner at Friends of the Earth, ‘the government says that some data centre growth has been factored into their new climate plan. But there are numerous proposals across the UK and if they all get the go-ahead it is difficult to see how they have properly balanced the carbon books’.

One thing is certain: the UK government’s mitigation strategy leans heavily on developments in major UK infrastructure. History tells us that takes time. The infamous High Speed Two (HS2) project was first discussed in 2008-2009. Parliamentary approval was confirmed when the High Speed Rail (London-West Midlands Act received Royal Assent in February 2017. Enabling works began in 2017-2018, only for the eastern leg to Leeds to be scrapped in 2021, and the Manchester leg cancelled in October 2023.

The HS2 programme covered a period of some 15 years without actually being properly constructed. Given 2050 is just 24 years away, the timings for AI power mitigation to be achieved might be considered, by some, to be tight.

The conclusion

It appears that it is realistic for the UK to be both an AI superpower and to meet its net-zero ambitions on time. However, it is unlikely to do so without heavily championing:

  1. Clean power expansion
  2. Energy-efficient data centre design and operation
  3. Better emissions measurement and policy alignment, with mandated reporting
  4. Smarter grid infrastructure and regulation
  5. Strong co-ordination between the technology and energy sectors

If the UK delays any of these efforts, or continues to rely on fossil fuels to meet data centre demand, it risks creating tension between its economic ambitions and its climate commitments. Proactive policy and technology solutions must be created now if the UK is to avoid the twin goals coming into conflict.

For more information

Contact the team