The discussion in the UK around AI data center development is shifting from investment scale to actual returns. A recent report by the economic think tank Verdant suggests that the long-term employment generated by such facilities may be far lower than previously anticipated by the tech industry.
By 2035, only around 10,400 new long-term jobs will be added.
Verdant estimates that the UK’s planned data centers will create approximately 10,400 long-term operational jobs by 2035—a figure roughly one-quarter of the 40,200 predicted by the UK technology industry group techUK.
The agency also stated that the current data centers in the UK directly support approximately 4,400 jobs, significantly lower than techUK’s estimate of 24,300.
Verdant believes that the UK government and industry stakeholders previously relied on widely cited industry data when supporting new data center projects, but these figures overestimated the number of jobs that would be permanently retained after project completion.
The disagreement centers on the number of jobs per megawatt.
The core of the dispute between both parties is the workforce scale corresponding to the unit capacity of the data center.
techUK's estimate, based on staffing and capacity levels typical of conventional data centers, arrives at approximately 8.5 jobs per megawatt. Verdant, however, argues that this metric more closely reflects earlier, smaller-scale data centers, which typically require more on-site personnel.
After reviewing planning applications and publicly available employment data for 20 proposed data centers in the UK, Verdant found that, on a capacity-weighted average, each megawatt corresponds to approximately 1.2 operational jobs.
The organization stated that the new generation of large data centers is more automated, resulting in fewer long-term permanent positions. It also noted that techUK’s higher estimate includes off-site roles, such as contractors and supply chain-related workers.
The government and industry stakeholders publicly refute
techUK responded that its data was based on industry information available at the time, that the methodology was transparent, and that it still reasonably reflects the industry’s contribution to employment. The organization also stated that Verdant’s sample size was too small and that it only counted jobs within data centers, ignoring the substantial external employment and service value supported by data centers.
The UK government also refuted the report. A government spokesperson said the report is misleading for using the metric of "jobs per megawatt" and incorrectly compares data centers to facilities such as hospitals and schools.
The speaker also stated that the report underestimates the broader economic value of computing infrastructure and overlooks the significance of the UK building critical digital infrastructure for national security and technological sovereignty.
The wave of AI development continues to advance.
Despite controversy over employment figures, the UK government continues to prioritize the expansion of AI infrastructure as a policy focus. The UK has previously designated data centers as critical national infrastructure, emphasizing their importance to public services and economic activity.
On the enterprise side, Alphabet, Microsoft, Meta, and Amazon have collectively committed nearly $700 billion this year to build dedicated AI data centers, reigniting market interest in technology roles related to data centers.
Randstad's previous analysis of 50 million job postings showed that demand for robotics technicians within data centers increased by 107% between 2022 and 2026; demand for cooling and HVAC system engineers increased by 67%; and demand for industrial automation technicians increased by 51%.
This means that AI data center construction will indeed create new job opportunities, but these are more likely to be concentrated during the construction phase, in supporting services, and in specialized technical roles, rather than in the large-scale, long-term on-site positions often promoted in industry marketing.