Wall Street’s biggest banks extend borrowing binge to fund AI boom

The biggest banks on Wall Street are borrowing at a record clip, and they don’t plan to slow down. The reason is artificial intelligence, or more precisely, the enormous pile of money it takes to build it.

Big Tech wants data centers, chips and power. Banks want to be the ones financing all of it. To do that, they first have to stock their own shelves with fresh capital.

According to analysis from Barclays Plc, the six largest US banks are poised to issue approximately $41 billion in senior debt in the fourth quarter of 2026. That would be a 30% increase over the long-term average for that stretch of the calendar.

The numbers behind the binge

The projected fourth-quarter haul follows a busy summer. Third-quarter issuance reached around $50 billion, more than double the amount from a year earlier.

Year-to-date global debt issuance from US banks now stands at $192 billion, up roughly 40% compared with the previous year. Full-year projections have been adjusted to $233 billion, with an even higher figure of $294 billion for US banks overall.

Senior debt sits near the front of the line if a borrower runs into trouble. It gets repaid before junior bonds and equity holders. For banks, selling it is a relatively straightforward way to raise funding and reinforce the balance sheet.

Why AI is doing the heavy lifting

The borrowing spree is closely tied to spending by the hyperscalers. That group includes Alphabet, Amazon, Meta and Microsoft, the companies running the giant cloud platforms that power much of the internet.

These firms are pouring historic sums into data centers, chips and the infrastructure that keeps them running. For years, they paid for that kind of expansion largely out of their own cash. That pattern has shifted toward heavy reliance on bonds and loans.

AI-related debt issuance in 2026 is somewhere between approximately $489 billion and $500 billion so far, with projections for the full year topping $570 billion.

Morgan Stanley, Goldman Sachs, JPMorgan, Bank of America and Citigroup are among the major banks involved in arranging these AI financings. Every bond they underwrite and every loan they syndicate generates fees. Every trade in the resulting securities feeds their trading revenue.

From cash-rich to leveraged

For most of the past decade, the largest tech companies were known for sitting on mountains of cash. That reputation is changing. Aggregate gross leverage among the hyperscalers has doubled recently, a sign of how much borrowing the AI buildout requires even for the richest companies on the planet.

That shift is precisely what has turned AI from a tech story into a credit story. When companies of this size move from self-funding to borrowing, the ripple effects reach bond markets, bank balance sheets and investors who never thought about data centers.

What this means for banks, tech and credit markets

The next checkpoint arrives soon. JPMorgan, Goldman Sachs and other major banks are expected to report earnings around mid-October 2026. Those results should offer a clearer view of how much the AI financing wave is actually showing up in revenue, and whether borrowing trends are likely to hold.

Rising leverage across the tech sector is one concern. Companies that once paid cash now carry meaningfully more debt, and the sustainability of those leverage levels is an open question if AI returns take longer to arrive than expected.

When banks and tech giants are both selling debt in large volumes, the market has to digest a lot of supply at once. If demand softens, borrowing costs could rise and financing conditions could tighten.

A large share of the activity flows through the same handful of banks and the same handful of borrowers. Watch how elevated debt levels affect valuations for both banks and tech firms. Watch whether bond spreads start to widen as supply keeps coming. And watch how bank executives describe the pipeline on their October earnings calls.