U.S. corporates have issued over 200 billion dollars of bonds this year specifically tied to artificial-intelligence (AI) infrastructure projects, marking a seismic shift in credit markets.

Among the most prominent deals, one company alone issued 30 billion dollars in bonds, drawing orders of roughly 125 billion dollars — the largest investment-grade corporate bond subscription on record.

Major issuers such as these are tapping debt markets after previously relying on strong earnings and balance sheets to fund AI data-centre build-outs.

Analysts estimate that AI-related issuance accounts for more than one-quarter of all net U.S. corporate debt supply this year.

The wave of issuance has prompted concerns among fixed-income investors.

Some warn that deploying huge volumes of debt into projects whose returns are still years away could elevate credit risk across sectors.

The longer-duration nature of these bonds also raises sensitivity to interest-rate shifts and market volatility.

Credit-market participants note that the influx of AI-linked debt may divert investor demand away from other corporate borrowers.

One fund manager described the trend as the “elephant in the room” for many investors, citing risks of concentration and capex sustainability.

Some issuance is also taking place through private-debt channels, adding opacity to the financing landscape.

While borrowing costs remain favourable and demand for corporate credit is strong, the scale and pace of these deals are unprecedented.

Many market watchers expect the trend to continue into 2026, intensifying the spotlight on how companies finance the next wave of AI investment.

With debt levels elevated and investment horizons long, the credit cycle may face fresh stress if returns on AI infrastructure disappoint or if a broader tightening of monetary policy reduces appetite for long-dated corporate bonds.