Who pays more to borrow when AI gets the money?

The business worth watching in an AI financing boom might be the one that has no use for a new chip. Imagine a manufacturer renewing a loan on a perfectly good production line. Does its lender now charge more because a much bigger borrower wants to finance a data center? At a November 3 investment summit in Hong Kong, Xiao Yuanqi, a senior official at China’s National Financial Regulatory Administration, argued that money flowing to technology leaders is driving up borrowing costs for traditional industries. That would leave businesses less room to reinvest and make their debt harder to service. Reuters reported his warning about global investment. It was not an announcement of a lending cap, and he did not put a number on what an otherwise comparable business is paying extra. The Bank for International Settlements offers a useful check. Its researchers traced how commitments from large technology firms can support debt issued by separate financing vehicles. Banks sometimes lend to those vehicles and to investors holding the debt. That puts the same promise behind more than one investment. Yet on days with AI-financing news in 2026, the effects the researchers measured in US Treasury and investment-grade corporate bond markets were small and short-lived. That finding neither proves nor disproves the Chinese official’s argument about traditional firms. A small business seeking a bank loan doesn’t refinance in the US Treasury market.