Two critical areas of UK Government policy, AI and the environment, may soon be on a very public collision course. One that could impact the communications and policy stances of companies, too.
The UK has legally committed to achieving net zero carbon emissions by 2050, and in parallel has a far-reaching AI policy agenda covering support for fast-growth businesses, responsible development, data stewardship and a pro-innovation regulatory agenda.
But, while it has long been understood that the creation of more high-power data centres will be critical to delivering on that AI vision, reports this week outlined that ministers are being briefed on the need to burn more fossil fuels - namely gas - to give the UK enough processing power in those data centres to be able to capitalise on the potential.
According to the Politico article, ministers have also been warned that the country’s comparatively high energy costs and length of time taken to access the national electricity grid risk standing in the way of energy-intensive data centre development and AI-related investment. The role of aligning governmental AI and climate ambitions rests with the AI Energy Council, which met for the first time recently, resulting in the leaked minutes that have caused the stir over a clash of ambitions.
The apparent need to call on gas to generate the surge in electricity capacity required signals the scale of the data centre power problem the UK faces, as well as likely being an attempt to set public expectations around concessions that may lie ahead.
Concerns have already been raised over AI-tied data centres not having to report their water usage, which the Environment Agency has called a “blind spot”.
The energy demands of AI processing is not a new story. But this is the first time that the sheer scale of power that the UK will need has been stacked up directly against the country’s energy transition intentions.
With the technological race to create advances and new value from AI very much a political priority, including the potential of AI to help to combat climate change, it’s doubtful that there will be any cooling of ambitions on that front.
But the prospect of a collision course with environmental goals is one that may see AI and its energy consumption impact connected more readily in the public eye, with more questions asked about how future electricity demands can be met - for data centres, but also for transportation, heating, air conditioning and other future drains on the grid.
