According to a recent report covered by Goldman Sachs, the surge in corporate borrowing to fund AI infrastructure — from data centers to hardware — is having uneven effects across credit markets. Investment‑grade (IG) bonds tied to AI‑heavy firms are seeing mostly issuer‑specific underperformance. Meanwhile, high‑yield (HY) bonds across the AI‑sector face broader, sector‑wide pressure.
In the IG segment, many AI‑linked bond baskets (excluding direct heavy AI issuers) have slightly outperformed broader markets in 2025 — suggesting that investors still see some value there when exposure is hedged and names are diversified. But bonds from firms directly financing massive AI‑driven investment — especially in data‑center buildouts — have under‑performed, indicating growing caution that their aggressive capex may not pay off.
The situation is worse in the HY market. High-yield AI‑linked debt — often issued by smaller firms or those with weaker balance sheets — is under pressure broadly, as investors grow skeptical about execution risk and long‑term viability. Credit analysts warn that a wave of similar speculative‑grade issuance from firms chasing AI infrastructure could further depress the segment, especially if demand slows or funding costs rise.
That said, some market players emphasize that the recent underperformance doesn’t mean widespread collapse. Many top-tier tech firms — those with strong cash flow and diversified businesses — remain in good standing, and a subset of IG AI‑exposure funds still look relatively stable. What’s changing is investor behaviour: risk‑aware allocation, selective exposure, and heightened scrutiny of companies carrying large, unproven AI‑infrastructure burdens.
