By Nate Wolf
Big Tech companies are breaking ground on data centers and making fresh chip deals on what seems like a daily basis. But we might only be scratching the surface of the artificial-intelligence spending boom.
In a report released Tuesday, J.P. Morgan boosted its estimates for AI-related capital expenditures to $5.5 trillion through 2030, up from a previous estimate of $5.1 trillion.
Spending will surge next year, the firm says. AI investments by the so-called hyperscalers -- Amazon.com, Microsoft, Alphabet, Meta Platforms, and Oracle -- will likely exceed $1.1 trillion in 2027, up from around $650 billion in 2026.
There are a couple reasons this quintet can't pull back now. For starters, data centers need a continual supply of advanced chips. While J.P. Morgan expects financing for data-center infrastructure to stabilize around 2028, financing for chips and custom AI accelerators will continue to surge into 2030.
Second, the transition from training AI models to deploying AI agents requires more compute capacity.
"We just don't have enough compute to allow for agentic adoption across the economy," the J.P. Morgan analysts wrote.
Companies are tapping every possible funding source to keep up with one another. The hyperscalers have issued around $170 billion in bonds this year, J.P. Morgan calculates. Some are now selling stock: Google-parent Alphabet recently issued $85 billion in equity to finance its AI buildout.
For now, the growing need to find and deploy capital isn't an issue.
"Hyperscalers remain remarkably profitable, with early returns on investment positive," the analysts wrote.
But the more money the hyperscalers invest, the more money they will have to make back to achieve acceptable returns for investors. Whether they can do so remains the multitrillion-dollar question.
If they fail, it won't be due to a lack of demand. Enterprise customers are already burning through AI budgets.
"Our biggest current fear is that adoption stalls as costs are too high to justify some use cases, until such time as additional compute can further reduce token costs," the J.P. Morgan analysts wrote.
In other words, Big Tech has no choice but to keep investing.
Write to Nate Wolf at nate.wolf@barrons.com
This content was created by Barron's, which is operated by Dow Jones & Co. Barron's is published independently from Dow Jones Newswires and The Wall Street Journal.
(END) Dow Jones Newswires
June 16, 2026 13:37 ET (17:37 GMT)
Copyright (c) 2026 Dow Jones & Company, Inc.
Comments