UK ministries clash over AI data‑centre power forecasts, leaving net‑zero timetable in limbo

The Department for Energy Security and Net Zero and the Department for Science, Innovation and Technology have presented mutually exclusive projections for the electricity that will be required to operate the United Kingdom’s rapidly expanding artificial‑intelligence data‑centre infrastructure, a discrepancy that emerged publicly in early April 2026 and immediately raised doubts about the coherence of governmental planning for both climate mitigation and digital leadership. The conflicting estimates, which differ by as much as a factor of two, have forced grid operators to contend with ambiguous demand curves at a time when renewable generation capacity is being scheduled on the basis of comparatively stable forecasts, thereby undermining the reliability of investment signals that underpin the nation’s net‑zero pathway.

According to the energy department’s internal modelling, the cumulative load from AI‑driven servers is expected to reach approximately 12 terawatt‑hours per year by 2030, a figure that would necessitate the commissioning of several gigawatts of new offshore wind and nuclear capacity, whereas the digital department’s more optimistic scenario projects a modest 6 terawatt‑hours annually, a discrepancy that would halve the required renewable infrastructure and consequently alter the projected emissions trajectory. Both ministries, however, have declined to produce a joint analysis or to subject their assumptions to an inter‑departmental review, citing divergent methodological approaches and the urgency of delivering separate policy milestones, an explanation that does little to mitigate the perception of institutional silos dictating national strategy.

The absence of a coordinated framework for estimating AI data‑centre power consumption has not only left the Department for Business and Trade to grapple with uncertain cost‑recovery mechanisms for the grid but has also compelled the Office for Climate Change to repeatedly revise its carbon‑budget calculations, thereby illustrating how unaligned departmental objectives can generate a cascade of administrative readjustments that erode confidence among investors and the public alike. Moreover, the lack of a unified stance on such a critical infrastructure component signals to industry that policy signals remain fragmented, potentially prompting AI firms to locate their high‑density computing facilities abroad where regulatory clarity and energy supply assurances are more readily available, a development that would undercut the very ambition of positioning the United Kingdom as a global AI superpower.

This episode, emblematic of a broader pattern in which climate‑related ministries and digital innovation bodies operate in parallel without a mechanism for reconciling their competing forecasts, underscores the systemic risk that arises when strategic objectives—namely decarbonisation and technological leadership—are pursued in isolation rather than through an integrated policy architecture designed to balance resource constraints with aspirational targets. Unless the government institutes a formal cross‑departmental coordination process that aligns energy demand modelling with digital growth strategies, future discrepancies of similar magnitude are likely to recur, perpetuating a predictable cycle of planning inertia that threatens both the credibility of the net‑zero commitment and the feasibility of the UK’s AI agenda.

Published: April 26, 2026