WEF Economists Reveal AI Boom's Catch: Data Centres Could Raise Power and Water Costs Without Big Jobs Boost

The AI boom is being sold as a jobs story. The people who model the global economy just described it as a power-and-water story that leaves the local hiring hall almost empty.

A World Economic Forum survey of chief economists found that 78% expect data-centre investment to make a significant contribution to global growth, while 61% do not expect it to drive a significant share of global job creation.

The economists also expect the build-out to put pressure on consumers. Some 78% expect data-centre expansion to materially increase average electricity prices paid by other consumers, while 58% expect higher average water prices for other consumers.

Separately, 79% expect significant pushback from local communities against new data centres.

The findings come from the WEF's Chief Economists' Outlook, published on 22 September 2026 after a survey conducted from 4 to 20 August 2026. The results are based on 36 survey responses and capture economists' expectations rather than a representative global poll or a site-by-site forecast of jobs, prices or investment.

Billions in Investment, Hundreds of Local Jobs

Evidence from the US suggests why the employment effect may be modest once construction ends.

A Brookings Institution analysis, originally published on 4 May 2026 and updated on 10 August, examined approximately 1,500 US data-centre facilities alongside 52 announced-but-cancelled projects. It found that a typical treated county gained roughly 100 to 200 jobs after receiving its first large data centre, depending on the type of facility.

Employment in data processing rose 56% over the first decade after a facility arrived, while telecommunications employment increased 43%. Wages were unchanged, while local home prices rose by about 2% to 5%. Construction can generate a much larger temporary workforce, but those jobs decline once a facility is built.

The scale of investment is nevertheless enormous. The WEF's Data Centre Sustainability and Resilience: A Decision Playbook for AI Infrastructure, published on 21 September 2026 with Oliver Wyman, projects global data-centre investment could reach $7 trillion by 2030, while data-centre electricity demand is expected to grow by at least 20% a year.

The report says the expansion will depend on surrounding power, water and infrastructure capacity, as well as the ability of developers to maintain support from host communities.

Higher Bills Could Meet Local Resistance

Data centres consume large amounts of electricity and, in many cooling systems, water. When new facilities arrive in areas with limited generation, grid capacity or water supplies, meeting that demand can require additional infrastructure whose costs may extend beyond the facility itself.

For other consumers, the WEF survey's expectation is specific: higher average electricity and water prices as data-centre demand expands. That cost exposure can reach households and businesses that receive none of the facility's direct employment.

The separate WEF playbook estimates that better sustainability and resilience practices could unlock between $700 billion and $1 trillion of planned global data-centre investment by 2030 by reducing delays and cancellations.

The Wider AI Jobs Market Is Sending a Different Signal

The broader AI labour market is showing a separate pressure point. Stanford University's Digital Economy Lab, in an update published on 12 August 2026, reported that employment among 22- to 25-year-olds in highly AI-exposed occupations was about 19% below the path implied by less-exposed peers, based on ADP payroll data through June 2026.

The researchers found the adjustment was concentrated in weaker hiring of younger workers, with no comparable decline among experienced employees, and no evidence of widespread economy-wide displacement.

Challenger, Gray & Christmas reported that employers cited AI as a reason for 101,743 announced US job cuts through June 2026. Those figures describe reasons given by employers, rather than verified cases in which an AI system directly replaced a worker.

The WEF figures point to a specific tension inside the AI build-out: investment and economic output can rise sharply without producing a comparable permanent payroll.

For communities hosting the infrastructure, the practical questions are how many jobs arrive, how long they last and who pays for the power, water and other systems needed to support them.