If you search the term “too big to fail” these days, the results aren’t all just about the Great Recession. The idea is popping up in relation to the artificial intelligence sector, a growing slice of the economy. Last year, big tech companies spent around $400 billion on data centers, outpacing consumer spending in the first half of 2025. In 2026, those expenditures are expected to grow. James van Geelen, founder and CEO of Citrini Research, spends a lot of time researching AI companies, which drove much of the market’s 2025 gains. Nvidia was up almost 40% last year, and Alphabet was up around 65%. “Even if the stock market were to go down, AI would still proceed as a technology,” van Geelen said, noting he doesn’t believe the technology will fail. “2026 is probably the year that we start seeing people losing their jobs and those jobs ceasing to exist,” van Geelen said. “That is scarier to me, from a sociological perspective, than being afraid that it’s not going to work.” But until artificial intelligence enters the stage where it really distorts the labor market, its economic impact is largely in the realm of capital expenditures and the data center buildout. One big reason AI is bringing up too big to fail now has to do with how it’s financed, said Patricia McCoy, a law professor at Boston College who has spent her career focused on government bailouts and ‘too big to fail.’“These companies, some of them, have borrowed financing,” McCoy said. “They’ve taken out loans and borrowed money through the bond markets.” Amazon, Meta, Alphabet, and Oracle all sold bonds last year to fund data center construction. Meta’s sold $30 billion in bonds in October, marking the single-biggest offering in the U.S. high-grade bond market last year. “The question is, will they be able to honor those debts, or will they default on them?” McCoy said. “That could trigger the failure of at least one large financial firm to whom it owes money, and the financial firm’s failure could set off a domino effect.” This domino effect is a marker of systemic risk — a key part of ‘too big to fail.’