UBS projected in a report covered on 21 October that AI infrastructure spending by major technology companies will exceed $1 trillion in 2027, after surpassing $600 billion in 2026, the Economic Times reported. The investment is adding to US economic growth, while placing greater demands on the companies’ cash flows.
Key points
- UBS expects AI-related capital expenditure by major technology companies to exceed $600 billion in 2026 and $1 trillion in 2027.
- The bank estimates that AI infrastructure spending added around 0.75 percentage points to US GDP growth in the first half of 2026.
- UBS says cash flows have weakened as investment has risen, while demand for the companies’ bond issuances has become less robust.
- Investors expect free cash flow to turn positive in 2028, provided the companies earn enough from AI services, according to UBS.
UBS sees spending growth peaking in 2027
UBS said demand for computing capacity had overtaken supply earlier than anticipated and that companies had increased their investment plans, The Hindu Business Line reported. The bank expects spending to continue rising over the next year. It said the rate of growth could peak in 2027 as the outlays weigh on cash flows and create a greater need for financing.
That distinction matters for the forecast. UBS projects a higher level of annual expenditure in 2027 than in 2026, while identifying 2027 as a possible high point for the pace at which spending grows. The bank tied that possibility to the cost of the investment and the funding required to maintain it.
The projection concerns capital expenditure on AI infrastructure by major technology companies. It is an estimate of what they will spend, rather than revenue they will earn from the capacity. AI Affairs has separately reported Gartner’s assessment that more than $3.5 trillion of AI infrastructure spending is at risk.
AI investment adds to US GDP growth
UBS estimates that AI infrastructure investment contributed around 0.75 percentage points to US GDP growth in the first half of 2026. The bank also said rising share prices for AI-linked companies had increased household wealth and supported consumption, the Economic Times reported.
The bank reported little clear evidence so far of a material rise in overall productivity or substantial job displacement attributable to AI. It estimated the effect on the overall unemployment rate at around 0.2 percentage points or less. Those economy-wide assessments sit alongside different employment changes in the sectors it examined.
Employment in selected high-technology industries, including software and computer systems design, fell 1.4 per cent year-on-year, according to UBS. Construction industries supporting data centre infrastructure recorded growth of nearly 3 per cent. The figures describe employment in those industries, while the bank’s estimate for unemployment concerns the broader labour market.
UBS also said AI demand was contributing to higher prices for software, digital goods and electricity. The bank said productivity gains from AI could, over time, reduce inflation.
Bond demand and the 2028 cash-flow test
Cash flows at major technology companies have weakened as AI infrastructure expenditure has increased, UBS said. The bank also reported less robust demand for their bond issuances. Its forecast for further spending therefore comes with a financing pressure already visible in the companies’ cash flows and in the market for their debt.
According to UBS, investors expect free cash flow to return to positive territory in 2028 if the companies generate sufficient revenue from AI services. That expectation places a revenue condition beside the spending forecast: the companies undertaking the investment must earn enough from the services it supports for cash generation to recover.
UBS said the investment cycle’s trajectory hinges on demand for computing capacity staying above supply and companies generating sustained revenue growth from their infrastructure outlays.
