Higher interest rates add to fears of AI bubble popping
Higher rates from the US Federal Reserve, the Bank of England, and other central banks will add increased pressure on the booming AI industry.
Speaking to HotMinute, Charles Lagares (Chief Operating Officer at EBC Financial Group) noted that rising global yields can fuel concerns about an AI bubble because AI valuations, the prices investors put on AI companies, rest heavily on profits expected years from now.
He pointed to the fact that AI-related and semiconductor shares have fallen sharply and doubts about whether AI earnings and spending can keep growing may have weighed on market sentiment since July.
“AI companies now pay for more of their spending with borrowed money, so rising interest rates hit them harder than they used to. The UK’s Financial Policy Committee (FPC) cited a Morgan Stanley estimate that global AI-related debt issuance had reached around $450 billion by early September, more than double the total for all of 2025,” he said.
“Interest on that debt has to be paid whether or not AI revenues arrive on schedule. The FPC warned that rising debt at AI firms, combined with limited transparency about the financing, could amplify losses if expectations disappoint.
“It also said that a reassessment of the productivity gains expected from AI could affect government bond markets as well as AI-related assets, because growth and public finance forecasts depend in part on those gains.”
Concerns about profitability
Since late 2025, Lagares noted that EBC has discussed the risks from soaring valuations, weak evidence of widespread profitability, and the concentration of gains in a few large companies.
“No one can say in advance whether AI prices are a bubble, meaning prices that have run far ahead of what the companies can earn, because that only becomes clear once prices fall.
“Debt gives an earlier warning, because losses grow when borrowers cannot pay and lenders pull back, and two signs would show it first.”
The first is whether AI revenues are growing fast enough to cover the interest on the money borrowed to build AI infrastructure. If they are not, borrowers may have to cut spending or sell assets, and lenders may take losses, he said.
The second is how much more interest lenders charge AI-linked borrowers than they charge safer ones. A wider gap would show that lenders see more risk, and new AI building would cost more to fund.
“Higher interest rates push both signs the wrong way, because debt costs more to carry and lenders ask for more in return.
“The FPC said AI financing could affect a wide range of investors and funding markets, which could include those that fund UK AI start-ups and data centres.”
