Three times. That is how much more private money the World Bank says it mobilized last year than it did three years earlier, a total of $112 billion, and it is the number Ajay Banga reached for on Tuesday when asked whether the artificial intelligence boom is crowding everyone else out of the credit market. His answer, delivered on CNBC's Squawk Box, was that the availability of capital is "not really an issue."
It is a confident claim, and it comes from a man whose institution sits on the receiving end of a great deal of global financing. That vantage point is worth keeping in mind.
A view from the development side
Banga, who took over the bank after a career that included running Mastercard and serving as vice chair at General Atlantic, did not deny that investors are nervous. He acknowledged, according to CNBC's account, that some are uneasy about circular financing arrangements and about the business model underpinning AI spending. Those worries are real in rich-world markets, where a handful of firms are committing enormous sums to data centers and chips, and where some of the money flows between companies that are also customers and suppliers of one another.
He drew a line, though, between that world and the one the World Bank works in. Emerging markets, he said, are not meeting the same hesitation, because the money they want goes to things like roads and bridges. Their appetite for AI tends toward smaller applications rather than the vast systems that preoccupy developed economies. Wealthy countries, in his telling, are looking for funding to support big use cases. Developing ones need something close to the opposite.
Tidy enough. Whether it holds up in a bond market that prices risk globally, and rarely respects borders so neatly, is another matter. Investors who hold emerging-market debt are usually the same institutions that hold American technology bonds, and when their appetite for risk contracts in one place it tends to contract in the other. The 2013 "taper tantrum," when talk of tighter American policy sent capital rushing out of developing economies that had done little to provoke it, remains the standard reminder that conditions set in the richest markets can be felt most acutely in the poorest.
Why the bank says its money is insulated
The second part of his argument concerned who gets the cash. The private capital the bank oversees, Banga said, does not go to governments. That matters because the sharpest fear around rising yields is sovereign: countries already stretched on debt service watching their borrowing costs climb as AI-hungry corporations compete for the same pool of savings. By steering private money elsewhere, he suggested, the institution sidesteps worries about looming debt crises and the strain of higher bond yields in some nations.
There is logic to it. Money routed to projects and companies is not a claim on a treasury. Yet the projects still sit inside economies whose currencies, sovereign ratings and central bank decisions shape what any borrower pays, and a road concession in a country whose government is being squeezed by yields doesn't float entirely free of that pressure. Rating agencies have long applied a sovereign ceiling to many domestic borrowers, which means a downgrade of the government can raise the price of private credit within its borders even when the project itself is sound. Banga didn't address the point in the portions of the interview that have been reported.
The rate decisions ahead
His remarks arrive just before a cluster of rate-setting meetings at central banks across the major economies. Policymakers and the economists who advise them have spent months trying to establish how broad adoption of AI will affect both the appetite for capital and the shape of the job market, and the evidence so far points in no single direction. If the technology draws in enormous investment while also reshaping employment, the effect on neutral interest rates, the level that neither stimulates nor restrains growth, is hard to call. Some argue upward, on the strength of investment demand. Others point to productivity gains that could eventually ease inflation.
Banga stayed on the optimistic side of that argument without claiming certainty. "I still believe that AI is going to be transformative and profitable," he said, before adding that everybody sees bumps in the road and nobody knows when the next one will arrive. It was a careful sentence, and probably the most honest one in the exchange. Anyone who says they can time the wobble is guessing.
What to watch
The World Bank, founded in 1944 to rebuild after the war and now tasked with ending poverty and supporting prosperity, has an institutional interest in telling the world that money is plentiful, and its mobilization figures are a measure of its own success. None of that makes Banga wrong. But the test lies elsewhere: in the spreads that developing-world borrowers actually pay in coming months, and in whether the central bank meetings now on the calendar nudge yields in a direction that turns his reassurance into an argument.
If rates keep climbing, the distinction between big AI and small AI, between sovereign and private, will be tested quickly. Capital may well be available. The open question is at what price, and for whom.