Global Banks' Direct AI Data Center Exposure Stays Modest

CreditSights says banks' direct lending to AI and data center projects remains relatively small, though indirect risks are harder to measure.

Updated on Nov. 17, 2026 in Data Centers

Global Banks' Direct AI Data Center Exposure Stays Modest

Live Poll

Do you trust your bank to be safe from risks tied to the AI boom?

CreditSights reported that global banks' direct exposure to AI and data centre financing remains relatively modest. The firm warned that indirect exposure through private credit funds and technology companies could be harder to quantify.

Why it matters

Banks are earning fee revenue from underwriting and advising on AI infrastructure debt, which makes the sector increasingly tied to bank balance sheets. CreditSights expects regulatory scrutiny to intensify as AI-related financing expands.

Direct exposure figures range from about 1.2% of euro area corporate exposures to roughly 3% to 5% of total loans at Singapore's three largest banks, with MUFG at 1.8% and SMFG at 1.9% of credit exposure. Indirect exposure through private credit and technology firms remains unquantified.

The players

CreditSights

CreditSights is a credit research firm that publishes analysis of corporate and financial-sector debt markets.

Mitsubishi UFJ Financial Group

Mitsubishi UFJ Financial Group is one of Japan's largest banking and financial services conglomerates.

Sumitomo Mitsui Financial Group

Sumitomo Mitsui Financial Group is a major Japanese banking group with global lending operations.

The details

CreditSights found that banks favour projects backed by investment-grade hyperscalers and long-term leasing agreements. Large North American and European banks are using significant risk transfer transactions to manage concentrated credit exposures and free up balance sheet capacity.

Timeline

  1. - November 17, 2026: CreditSights publishes its report on global bank AI exposure.

The Tech Race

Banks are financing AI infrastructure as hyperscalers expand data centre capacity, positioning lenders alongside the technology firms competing to build it. Their exposure remains concentrated in investment-grade borrowers and leased facilities, which limits direct risk while the indirect links through private credit grow harder to track.

Because banks are financing data centres through debt and fee-generating deals, the cost and availability of that credit can shape the pace of AI services that users rely on. Indirect exposure through private credit funds may also affect the investment products many savers hold.

The takeaway

Watch how banks disclose exposure through private credit and technology partners, since that is where the largest unmeasured risk may sit. Investors should review fund holdings for indirect ties to AI infrastructure debt.

Further reading

For more on how AI infrastructure is being built and financed, see our Data Centers coverage.

Source note: This article includes information reported by BusinessToday.

Live Poll

Do you trust your bank to be safe from risks tied to the AI boom?