Strategists at Goldman Sachs Group forecast that the five largest U.S. hyperscale cloud service providers will increase their spending on AI infrastructure by more than 50% next year, reaching $1.2 trillion.

A team led by Ryan Hammond stated in a report that this figure exceeds the Wall Street consensus estimate of $1.1 trillion in capital expenditures for 2027.

The team noted that for these investments to break even, these companies would need to generate approximately $300 billion in AI revenue annually over the coming years.

Hammond and colleagues said, "Based on market consensus, capital expenditures as a share of GDP in 2027 are set to surpass any tech investment cycle since the large-scale railway construction of the late 19th century."

Amazon, Alphabet Inc., Microsoft, Oracle, and Meta Platforms Inc. are expected to invest $800 billion this year to build AI capabilities.

The Goldman Sachs team indicated that although the scale of spending will increase substantially next year, the pace of capital expenditure growth is expected to slow.

The strategists wrote, "Both the growth rate of capital expenditures and the extent to which actual spending exceeds market consensus will weaken compared to recent quarters."

Currently, hyperscaler capital expenditures have already surpassed operating cash flow, making it increasingly necessary to raise funds through debt and equity financing.

Potential constraints include shortages of power, labor, and memory chips, as well as restrictions on data center construction.

Goldman Sachs strategists expect hyperscaler capital expenditure growth to slow from nearly 100% this year to 54% in 2027, and further decelerate to 12% in 2028, by which time capital expenditures are projected to reach $1.4 trillion.