AI Debt Boom Puts Markets and Governments Under Strain

A multitrillion-dollar AI spending surge has helped push stocks to record highs. But debt behind the boom is now lifting borrowing costs and exposing risks across global markets. Washington’s light-touch approach to business helped fuel the investment, while ordinary people face the consequences as inflation, energy shocks and government debt leave less room for economic decisions.

“Rising rates always seem to sniff out and expose the most vulnerable time after time,” said Jim O'Neill, former Goldman Sachs chief economist and former British Treasury minister. Countries and businesses with greater exposure suffer most, he said. The vulnerability extends beyond firms and investors betting on artificial intelligence.

Who Carries the Risk

Investors grew concerned as U.S. President Donald Trump's trade policies and the Iran war's energy shock revived inflation and pressured central banks to raise interest rates. AI capital investment and heavily indebted governments also pushed borrowing costs to multidecade highs. U.S. midterm elections, volatile geopolitics and central-bank policy added uncertainty in the year's final stretch.

The strain is already showing up in the cost of essentials and the limits governments face. Oil reached $105 a barrel, up almost 75% so far that year, after the Iran war disrupted supplies for months. Some emerging-market currencies, including those in Indonesia and the Philippines, hit record lows as higher crude prices forced central banks to raise rates.

In France, bond yields rose to levels last seen during the 2012 euro zone debt crisis, limiting the government's room to respond to protesters during what the report called the worst civil unrest in years. Japanese bond yields recently reached 30-year peaks as the energy-import-dependent economy faced pressure to raise rates faster, potentially undermining Tokyo's spending goals and disrupting global capital flows.

October delivered a sharp jolt to the AI trade. Investors shifted out of AI stocks and into safe-haven bonds, wiping more than $2 trillion in market value from the tech-heavy Nasdaq 100 last month. Big Tech earnings intensified scrutiny of whether those investments could generate returns quickly enough. Proposed data-center projects struggled to secure funding, while technology stocks in South Korea, China and the United States fell.

A Boom Built on Borrowing

Danske Bank estimated that half a dozen U.S. technology firms known as hyperscalers were poised to invest $1.2 trillion in AI infrastructure next year, on top of an estimated $900 billion they were forecast to spend this year. That planned investment equals as much as all European companies combined. Technology and AI stocks made up more than 40% of the S&P 500 and drove much of the index's 70% rise over five years.

The S&P 500 had fallen 3% from its early-October peak as investors pulled back from AI. The rate-sensitive Russell 2000 stood nearly 8% below its August peak. On October 22, the benchmark 10-year Treasury yield, which guides borrowing costs globally, dropped almost 20 basis points, its largest daily fall since early 2023. It remained around 5%, a level some investors view as a risk to broader financial stability.

Sheila Bair, who chaired the Federal Deposit Insurance Corporation during the 2008 crisis, pointed to rapid lending growth in parts of the financial sector, where higher rates could threaten business models. “You have a buildup of leverage and AI companies that are relying a lot more on debt now to finance their ambitions,” Bair said. “I also worry about the lack of fiscal firepower if we have another crisis because we're just so strapped.”

Power Without a Safety Net

Global debt totaled more than $350 trillion, a record and almost three times the size of the global economy, according to the Institute of International Finance. In its July financial stability report, the Bank of England warned that global shocks could interact and magnify their combined impact. Governments burdened by debt would have less capacity to address the economic and social effects of a major downturn.

Some analysts said the U.S. economy might already have slipped into recession without AI spending, dragging other economies with it amid the Iran war's energy-price surge. U.S. economist Paul Krugman cited persistently high inflation, rising government bond yields and cracks in the AI boom as major concerns. “There are troubling aspects of the current economy that could potentially be harbingers of the crisis to come,” Krugman wrote in an October 25 column on Substack.

The report also offered reasons for optimism. Recession warnings over the previous two years had proved wrong, Europe's economic data had been more resilient than expected, labor markets had remained stable, and 84% of S&P 500 companies reporting third-quarter results had beaten consensus estimates, according to LSEG data.

White House spokesperson Matthew Middleton said the Trump administration was committed to maintaining U.S. leadership in AI, fueling revenue growth and driving productivity across the economy. The administration's statement stands alongside the debt, rate and market risks that have accumulated around that investment. Tina Fordham, founder and geopolitical strategist at Fordham Global Foresight, said divisions among traditional allies could make it harder for major economies to coordinate rescue packages and monetary stimulus as they did during the 2008 global financial crisis. “The tide is AI,” she said.