Digital Infrastructure's Double-Edged Promise: UK Data Centres Under Job Creation Scrutiny

The UK’s ambitious push for new AI data centres, championed by the tech sector and government as a jobs engine, is facing a stark reality check. New analysis from environmental thinktank Verdant suggests that the projected economic benefits, particularly in job creation, have been drastically overstated, challenging the 'build first, ask questions later' approach to critical digital infrastructure.

Verdant's research directly refutes the industry lobby group TechUK's claims of over 40,000 new jobs from planned data centre facilities. Instead, Verdant predicts a significantly lower figure: just 10,400 direct employees. This represents a mere quarter of the industry's optimistic projections, a discrepancy that raises serious questions about the basis for rapid infrastructure rollout, which has seen Labour defend the initiative amidst growing public scepticism.

The divergence in figures is not just about raw numbers. Verdant's analysis, drawing on publicly available staffing data from 20 projects, reveals a critical inefficiency: existing data centres currently create approximately 8.6 jobs for each megawatt of energy consumed. This stands in sharp contrast to TechUK’s projection of 43.7 jobs per megawatt. Even more concerning, the thinktank posits that future projects, often on a far larger scale, are likely to be even less job-intensive, potentially yielding as few as 1.6 permanent positions per megawatt. This focus on permanent operational roles, distinct from temporary construction jobs, highlights the core issue of long-term economic contribution.

The resource intensity of these vast data warehouses further complicates the narrative. Verdant explicitly compared the 8.6 jobs per megawatt from data centres to approximately 400 jobs in a car plant and more than 3,600 in a hospital or health facility. This dramatic comparison underscores the comparatively low direct employment yield against significant energy and water demands, potentially conflicting with broader climate ambitions and crowding out alternative plans for electricity capacity. While the government rejected this comparison, questioning Verdant’s methodology and citing the 'wider economic value of compute' and 'national security,' the underlying challenge to the economic justification remains potent.

This debate transcends mere statistical squabbles; it signals a critical need for rigorous, independent evaluation of large-scale infrastructure projects. When government bodies echo industry claims without robust third-party verification, particularly for initiatives classified as critical, it risks misallocating resources and public trust. The implications extend to any economy contemplating significant investment in similar digital infrastructure: the allure of 'AI and related technologies' must be weighed against tangible, verifiable economic and environmental costs, rather than relying on figures that independent analysis consistently finds to be 'deeply misleading.' The case of UK data centres serves as a sharp reminder that the digital revolution, while promising, must be grounded in transparent and verifiable economic benefits.