JPMorgan forecasts $870 billion in AI capital spending by 2026, with a shift toward commercialization.

J.P. Morgan’s July 20 report forecasts global AI-related capital expenditures to reach nearly $870 billion in 2026, a 77% year-over-year increase, with hyperscalers accounting for approximately $750 billion. In 2027, Google’s capital expenditures are expected to be around $300 billion, Amazon’s $300 billion, and Meta’s $200 billion. The five major tech giants’ bond financing is projected to rise from $40–50 billion in 2022 to approximately $190 billion in 2026; Google completed an $85 billion equity financing in June. J.P. Morgan believes the AI investment thesis is shifting from capital expenditure competition to commercial monetization, with hyperscalers offering a more favorable risk-reward profile than AI hardware over the medium term. Semiconductor revenue excluding memory is expected to grow by over 30% this year, but the hardware sector has recently corrected by approximately 15%. Google, Microsoft, Meta, Amazon, and NVIDIA are set to release earnings reports soon, with capital expenditure guidance serving as the key indicator.