Fed’s Warsh warns AI investment is not immune to higher interest rates

Federal Reserve Chair Kevin Warsh rejected the idea that AI’s expected rewards make borrowing costs irrelevant. Speaking after the Fed held its benchmark rate at 3.75%–4% on October 28, he said interest rates still reach AI investment—a warning that even an ambitious technology buildout remains tied to financing conditions.

Warsh’s argument was not that higher rates would stop AI spending outright. He pushed back against a narrower claim: that unusually large expected returns would leave the sector largely untouched by monetary policy. “I don’t believe that they’re somehow divorced from the rest of the economy,” he said, according to Fortune.

A financing filter, not an off switch

Higher financing costs can influence which AI projects receive money, how much they attract and the returns investors demand. For data centers and computing infrastructure, those costs can weaken a project’s economics even when investors still find its long-term opportunity attractive. They can also make execution risks harder to overlook.

Warsh pointed to widening credit spreads for some assets as evidence that investors were becoming more selective. Yahoo Finance reported that he also described tighter financial conditions since September’s rate increase, with investors drawing distinctions between technology firms and other companies. His evidence concerned how financing was being allocated, rather than a claim that AI investment had stopped growing.

Interest rates get in the cracks. Interest rates affect directly, or indirectly, almost everything.

Kevin Warsh, Federal Reserve chair, at the October 28 press conference

The counterargument: profits can outweigh borrowing costs

The opposing argument rests on the scale of the prize. Pershing Square CEO Bill Ackman questioned in a September post on X whether higher rates would reduce AI demand and investment at all. As Yahoo Finance recounts, he argued that winning the race for superintelligence could offer a “near-infinite ROI,” or return on investment, making conventional economic models less useful.

Joe Brusuelas, chief economist at RSM, pointed to third-quarter investment figures as a challenge to the idea that rates were restraining the boom. Private nonresidential investment rose nearly 13%, while equipment outlays increased more than 20%. Brusuelas said the former likely included data centers; those figures were not a standalone measure of AI spending.

Moody’s chief economist Mark Zandi offered a conditional version of that argument. He told Yahoo Finance that high expectations for future profits could allow the largest cloud companies to absorb current interest rates and considerably more. In his view, AI investment could remain resilient as long as those expectations held.

But Zandi also warned that higher rates could raise the cost of capital and weigh on stocks. If investors reassessed AI companies and markets corrected, that could help trigger a recession. He described that outcome as a risk, not his baseline forecast—leaving room for resilient spending and financial vulnerability to coexist.

The Fed has reasons to wait

The central bank kept rates steady while seeking more economic data, with inflation elevated and growth and business investment strong. Warsh cautioned against assigning all that strength to technology: high-tech investment was leading the acceleration, he said, but spending was strengthening elsewhere, too. The policy decision therefore reached beyond the AI financing debate.

Julius Bendikas, U.S. head of economics and dynamic asset allocation at Marsh, told Fortune that balanced labor markets, subdued wage growth and anchored longer-term inflation expectations supported patience. He expected Treasury yields to have room to fall as expectations for future hikes moderated. Yet he also flagged strong growth, elevated energy prices and large fiscal deficits as risks that could push yields higher.