Is artificial intelligence compatible with net-zero?
5 min read
AI will revolutionise the state. But how will we power it? Ethan Dodds, Political Campaign Manager at Total Politics, suggests we must stop treating AI and energy policy as competing priorities
In an ageing society, the demands on public services will only grow, just as the pool of working-age taxpayers shrinks. If the state is to survive this demographic shift, it must do more with less.
That is the key offer of AI. From accelerating cancer diagnoses in the NHS to freeing up frontline police time and automating fraud detection, the benefits of AI are already making a difference.
But a narrative has emerged in Westminster that threatens to stall this progress. Seizing the benefits of this technology requires immense compute capacity, and the data centres that provide it require a connection to an already gridlocked electricity network. This has created a new dividing line in UK politics: do we prioritise data centres or renewable energy projects for a grid connection?
This is the wrong debate. Britain can, and must, pursue both.
New renewables will provide the clean, sovereign energy needed to kickstart the AI boom that will transform public services, while AI will help manage an increasingly complex modern grid, ease demand spikes and get shovels in the ground more quickly for new renewables. The net-zero transition and the tech revolution are not competing priorities, but two sides of the same coin.
A grid fit for the future
Britain’s energy architecture was designed for a long-gone era, built to manage a stable, predictable baseload of electricity from coal and gas. Today’s system is greener but vastly more complex. Wind and solar power fluctuate with the weather, creating mismatches between where power is generated and where it is needed.
Human operators, however skilled, cannot manage this complexity alone, and certainly not with the legacy technology currently at their disposal. Here, AI could be game-changing. Machine learning systems can predict demand hours in advance, forecast drops in renewable generation before they occur and identify transmission bottlenecks before they cause outages.
The net-zero transition and the tech revolution are not competing priorities, but two sides of the same coin
It can also help reduce demand spikes to ensure there is enough clean power to go around. By encouraging businesses and households to shift consumption away from peak periods, we can reduce our reliance on expensive, polluting gas-fired power stations that are switched on when energy from renewables is unable to meet demand. The result will be lower bills for consumers and a lower carbon footprint.
If we want an energy system that is both clean and reliable, we must modernise our infrastructure. By harnessing the benefits of technologies like AI, we can transform our grid from a relic of the nineteenth century into a system fit for the twenty-first.
Unblocking the connections queue
The notorious 15-year connection delays plaguing the grid are not just a problem for data centres. They also threaten the transition to clean power.
To hit the government’s 2030 clean power targets, renewables need to come online at a pace Britain has consistently failed to manage. Right now, the connections queue is the single biggest obstacle – clogged up, in part, by speculative applications from ‘phantom’ projects that have no serious prospect of ever being built. Automated screening systems that rapidly audit the queue and flag unviable applications will free up space for genuinely shovel-ready infrastructure.
Furthermore, by harnessing digital twins – virtual replicas of physical grid infrastructure – network operators can stress-test the system under different scenarios before a single brick is laid. This would dramatically shorten planning timelines, reduce cost overruns and give investors the certainty they need to commit funding to the infrastructure Britain desperately needs.
The wider carbon dividend
Critics of AI tend to concentrate on what data centres consume. That is the wrong focus.
The International Energy Agency has pointed to AI as one of the most significant tools available for cutting emissions across heavy industry, transport and commercial buildings – sectors that account for far more global carbon than data centres themselves.
Across these areas, AI is already delivering measurable gains. Smart monitoring systems are finding thermal inefficiencies in manufacturing plants. AI-managed traffic systems are cutting congestion, reducing idle time and fuel waste. And in agriculture, AI tools are optimising fertiliser use and lowering associated emissions.
Taken together, these efficiency gains are likely to outweigh the electricity consumed by the data centres that enable them. Restricting compute capacity in the name of carbon reduction would therefore be a strategic mistake. It would limit the deployment of the most powerful decarbonisation tools available to the UK’s highest-emitting sectors, while doing little to address the underlying constraints of an outdated energy grid.
Future-proofing the net-zero transition
If Britain wants a clean energy system that works, it must embrace the technologies that make it possible. If we stick with the status quo, renewable energy projects will take longer to build, the grid will continue to fail, and already squeezed households will pay the price through higher energy bills.
The net-zero transition and the AI revolution are, at their core, the same project: building an economy powered by abundant, clean electricity and the smart systems capable of managing it. The countries that recognise that first and align their industrial strategies accordingly will reap the rewards. Will Britain be one of them?
