UBS projects that major technology firms will spend more than $1 trillion on AI infrastructure in 2027, a figure that dwarfs the estimated $600 billion outlay for 2026. The bank’s report, released this week, attributes this surge to demand for computing power that has outpaced supply ahead of schedule.
This capital injection is no longer just a corporate budget line item; it is a macroeconomic engine. According to UBS, AI infrastructure spending added 0.75 percentage points to US GDP growth in the first half of 2026. Rising equity prices linked to these firms have also inflated household wealth, creating a secondary boost to consumer spending.
Yet the money is coming from somewhere, and the strain is visible. The report flags weakening cash flows among these tech giants as they race to expand data centers. Bond issuance demand has softened, and investors are now betting on a return to positive free cash flow by 2028—a timeline that hinges entirely on whether AI services generate enough revenue to justify the upfront costs.
The labor market tells a mixed story. While software and computer systems design saw a 1.4% year-on-year employment decline, construction industries supporting data center builds grew by nearly 3%. UBS noted that the impact on overall unemployment remains limited, at roughly 0.2 percentage points or less.
The next milestone is clear: whether demand for compute continues to exceed capacity. If the tech giants can convert this heavy infrastructure spend into sustainable revenue, the cycle continues. If not, the cash flow pressure could force a sharp correction in investment pace by 2028.