JPMorgan lifts AI spending forecast to $5.5 trillion through 2030

JPMorgan Global Research has raised its forecast for global AI-related capital spending to $5.5 trillion through 2030, according to its midyear outlook reported by Fortune. The estimate matters because it suggests Wall Street still sees room for the AI infrastructure boom to run, even as investors question whether demand can catch up with the scale of the buildout.

The bank previously estimated $5.1 trillion in AI-linked capital expenditures. JPMorgan attributed the higher projection to larger capacity expansion and a greater role for borrowing in funding data centers, chips and related infrastructure, Fortune reported.

JPMorgan’s outlook centers on what it calls the “AI upstream” buildout: the physical and computing backbone needed to support artificial intelligence. That includes data centers, semiconductors and supporting systems, with the U.S. still the main hub for investment.

According to JPMorgan, the U.S. accounts for about 85% of AI and machine-learning venture capital. The bank also expects benefits to reach China, South Korea and Taiwan because of their positions in semiconductor supply chains.

Debt is taking a larger role

JPMorgan increased its estimate for debt financing tied to the AI buildout to $4.1 trillion, Fortune reported. The bank cited higher loan-to-cost ratios, a measure of how much of a project’s cost is covered by borrowing.

Those ratios average above 85%, and some top 90%, according to JPMorgan. The bank said that reflects favorable credit conditions and investors’ belief that AI infrastructure projects are creating value.

Fortune reported one example cited in the outlook: a $15 million-per-megawatt investment can translate into a $25 million increase in market capitalization in some cases. JPMorgan said recent equity issuance is also strengthening company balance sheets.

Analysts cited by Fortune said higher leverage appears to be a strategic decision by companies using debt while markets remain receptive. Those companies are preserving the option to reduce leverage later, according to the report.

Hyperscalers still have cash flow

JPMorgan said hyperscalers are driving much of the AI infrastructure spending and are entering the current phase from a strong financial position. The bank expects their capital expenditures to reach $650 billion in 2026 and exceed $1.1 trillion in 2027.

Despite that spending, JPMorgan projects operating cash flow for hyperscalers will top $900 billion by 2027, Fortune reported. That projection underpins the bank’s view that the AI capital expenditure cycle remains profitable for now.

The spending remains concentrated among a small number of large technology companies, according to JPMorgan. The bank warned that a pullback by those firms could have a large effect on industry growth.

Cloud providers including Amazon, Google and Microsoft have reported rising AI revenue, Fortune reported. Investors remain divided over how long it will take for returns from AI products and services to justify the capital being committed.

Credit markets are central

JPMorgan expects high-grade corporate debt to provide more than $2.1 trillion of data center financing over five years. For 2026, the bank forecasts $150 billion of U.S. hyperscaler issuance and another $100 billion equivalent outside the U.S.

The bank also expects about $170 billion in financing from data center and chip issuers outside the core high-grade market, according to Fortune. JPMorgan said those alternative channels will remain smaller than the main corporate debt market.

JPMorgan’s conclusion is that AI capital spending is expanding quickly while the economics remain sound, at least at present. The open test is whether AI adoption grows fast enough to support the trillions of dollars now being directed into capacity.

This story draws on original reporting from Fortune.