Lenders Are Getting Pickier About Financing the AI Build-Out

Lenders and investors are demanding stricter terms to finance the AI build-out, according to The Wall Street Journal, which published its report on October 31, 2026. Chips, power and land are still constraints, but the price and conditions of the money paying for them now look like the tighter bottleneck.

The volume of projects needing funding is large and growing fast. Reuters reported, citing John McWilliams, that U.S. data center capacity under construction rose 48% between December 2025 and June 2026, measured in gigawatts (a gigawatt is one billion watts, roughly the output of a large power plant). Each gigawatt needs a financing stack: equity from owners, bank loans, bonds and sometimes private credit, which is lending from non-bank firms. Each layer is now being assembled under tighter scrutiny. The Wall Street Journal has described the build-out as on track to become the biggest economic bet in U.S. history.

Pricing is the most visible signal. Bloomberg reported that at the start of 2026, the highest-rated data center bonds offered yields below 5%. Yield is the annual return a bond pays measured against its price, so a rising yield means investors want more compensation for the same paper. That starting point is useful for judging what comes next.

Lenders are also pricing in the physical side of the build. Reuters Breakingviews reported that lenders are charging more and demanding tougher terms to protect against supply-chain bottlenecks, the kind that delay equipment and push back completion dates. A late project services its debt later. Senior bankers told Reuters they now scrutinize community concerns when assessing U.S. data center project loans, which adds a permitting and reputational dimension to credit work that lenders have not had to model at this scale.

Underwriting, the process of judging whether a borrower can repay and whether collateral (the asset pledged to secure a loan) will hold its value, has become harder for a more basic reason. Bloomberg reported that investors are struggling to underwrite data center deals because requirements for cooling and computing density are changing rapidly. Think of a mortgage on a house that is being redesigned while the bank decides whether to lend. If the building's specification shifts during financing, the asset is harder to value on the day the loan is made and in the years after. That is a different problem from a weak borrower, and a higher spread alone, the extra interest charged above a benchmark rate, does not solve it.

Some of the exposure may not yet be visible. S&P Global estimates, as reported by Reuters, that the world's biggest tech firms likely have more AI-linked liabilities than currently known. Off-balance-sheet structures, leases and special-purpose vehicles, which are separate legal entities set up to hold assets or debt, are the usual places such obligations sit. Analysts who rely on reported debt figures may be working from an incomplete picture.

The leveraged finance market, which lends to borrowers carrying high debt relative to their earnings, is also part of the story. Bloomberg reported that Morgan Stanley expects about $20 billion of AI-related deals in leveraged finance markets in 2026.

The open question, to my mind, is whether stricter terms slow the build or mainly reprice it. If lenders demand higher spreads, tighter covenants (promises that protect the lender if a borrower's finances weaken) and more equity, projects with thin sponsor support or uncertain tenants will struggle to close, and smaller developers will feel it first. If the largest sponsors can still raise capital on workable terms, the adjustment lands on smaller players and on the cost of each additional gigawatt. The reporting so far points to repricing and selectivity more than a halt. The distinction still matters for anyone holding this debt, underwriting it or depending on the capacity it finances.

For ordinary savers and borrowers, the connection runs through the bonds and loans held in pensions, insurance portfolios and mutual funds. Those holdings are where any change in the cost of AI infrastructure credit would eventually surface.