1. PIMCO AI debt warning centers on the long end of the bond market
PIMCO's Lotfi Karoui says AI-related borrowers accounted for nearly half of year-to-date gross supply of non-financial US investment-grade corporate bonds maturing in more than 15 years, as of late October 2026. The PIMCO AI debt analysis argues that portfolios built only on corporate bonds are less diversified than they appear and recommends looking elsewhere for balance.
Karoui, a managing director and multi-asset strategist at PIMCO, contends that issuance tied to artificial intelligence now carries outsized weight in the broader debt market. The concentration is most visible in the longest maturities, where companies building AI infrastructure have become dominant sellers of new bonds.
The analysis shifts attention from the stock market, where the AI boom has been most visible, to the credit market, where much of the financing for data centers and computing capacity is increasingly being raised. It also raises a practical question for bond investors: how much AI exposure they may be carrying without realizing it.
2. What is PIMCO saying about AI debt?
PIMCO's core message is that AI borrowing has become so large in the long-dated segment of the investment-grade market that it distorts the risk profile of corporate bond portfolios. When nearly half of new long-term non-financial supply comes from one theme, investors who buy that market broadly end up with a concentrated bet on that theme.
Investment-grade bonds are debt securities issued by companies with relatively strong credit ratings, meaning rating agencies judge them to have a lower risk of default. They are a core holding for pension funds, insurers and many individual investors seeking steady income. The non-financial segment excludes banks and other financial firms, covering industrial, technology and other operating companies.
Long-dated bonds, those maturing in 15 years or more, are particularly sensitive to changes in interest rates and to the long-term prospects of the issuer. A concentration of AI-related borrowers in this part of the market means investors are making a decades-long commitment to the success of the infrastructure those bonds finance.
3. Who is doing the borrowing
PIMCO attributes the surge mainly to hyperscalers and data center operators that are expanding AI infrastructure. Hyperscalers are the largest cloud computing providers, companies that operate vast networks of data centers and rent server capacity to businesses, developers and governments.
These borrowers have shown a strong preference for long maturities. In certain periods, hyperscalers have represented more than 30% of issuance in the 30-year segment, compared with roughly 13% of 10-year bonds. That skew toward the longest end of the curve explains why the concentration is so pronounced among bonds maturing beyond 15 years.
Long-term borrowing is a natural fit for projects such as data centers, which are expensive to build and are expected to generate returns over many years. Matching the life of the debt to the life of the asset is a standard approach in corporate finance. The consequence, however, is that bondholders take on exposure to how those assets perform over a very long horizon.
4. Why concentration matters for bondholders
Diversification is one of the main reasons investors hold bonds alongside stocks. A well-diversified bond portfolio spreads risk across many issuers, industries and economic drivers, so that trouble in one area does not dominate the overall result.
When a single theme accounts for a large share of new issuance, that protection weakens. An investor who buys a broad index of long-dated investment-grade corporate bonds may assume they own a diverse slice of the US economy. In practice, a growing portion of that exposure is tied to the same AI infrastructure build-out that is already a major driver of equity markets.
This overlap means that a setback for AI investment could hit both a portfolio's stocks and its bonds at the same time, reducing the cushioning effect that bonds are meant to provide. PIMCO's point is not that AI borrowers are weak credits, but that too much of the market now moves with a single story.
5. Why long maturities amplify the exposure
The concentration is especially significant because of where it sits on the maturity curve. Bonds that mature in 15, 20 or 30 years behave differently from shorter debt, and those differences magnify the effect of any single theme that dominates their supply.
Duration, a measure of how sensitive a bond's price is to changes in interest rates, rises with maturity. A 30-year bond can lose or gain far more value than a five-year bond when rates move by the same amount. Investors who hold long-dated debt therefore accept larger price swings in exchange for locking in income over a longer period.
Long maturities also stretch the period over which an issuer's business must remain healthy enough to keep paying interest and eventually return the principal. For AI infrastructure, that means bondholders are relying on demand for data center capacity, and on the economics of the technology it supports, holding up for decades rather than a few years.
Pension funds and insurance companies are among the most important buyers of long-dated corporate bonds, because their obligations to retirees and policyholders also stretch far into the future. They match long-term liabilities with long-term assets. As AI-related borrowers take up a larger share of that supply, these institutions may find it harder to build long-dated portfolios that are not heavily tilted toward a single industry.
This is the context in which PIMCO's recommendation should be read. The firm is not calling for investors to abandon long-term corporate credit, but to recognize that the composition of that market has changed and to rebalance accordingly using assets with different underlying drivers.
6. Where PIMCO suggests investors look instead
Karoui recommends diversifying away from AI-heavy corporate credit without giving up credit quality. The alternatives PIMCO highlights are agency mortgage-backed securities, higher-rated non-agency residential mortgage bonds and emerging-market debt.
Agency mortgage-backed securities are bonds backed by pools of home loans and linked to US government-sponsored housing agencies. Because of that connection, they are generally regarded as carrying very low credit risk. Non-agency residential mortgage-backed securities are private-label equivalents issued without that agency link, and PIMCO favors the higher-rated portions of that market.
Emerging-market debt refers to bonds issued by governments and companies in developing economies. These securities respond to a different set of economic and political forces than US corporate bonds, which can make them useful as a source of diversification.
7. The case for mortgages and emerging markets
PIMCO's argument for these alternatives rests on two factors: lower correlation with AI-heavy corporate bonds and better carry. Correlation describes how closely two investments move together; assets with low correlation tend not to rise and fall at the same time, which helps smooth a portfolio's overall returns.
Carry is the income an investor earns simply by holding a bond, mainly through its interest payments. Better carry means a higher yield for a given level of risk, allowing investors to diversify without sacrificing income.
PIMCO also views emerging-market debt and high-quality securitized products as offering better risk diversification and attractive yields relative to AI-dominated corporate investment-grade bonds. In other words, the firm sees a way to maintain portfolio quality and income while reducing dependence on a single sector.
8. Political risk around data centers
The analysis also points to a political dimension. The results of the November 2026 US midterm elections could create challenges for data center development, as proposed restrictions may delay new projects.
PIMCO believes such restrictions may slow the pace of data center construction and the related bond issuance. Data centers require large amounts of land, electricity and water, and they have become a subject of local and national debate in a number of places, which makes them sensitive to changes in policy.
For bond investors, any slowdown in issuance would affect the supply side of the market, potentially easing the concentration PIMCO describes. It would not, however, change the exposure already embedded in portfolios that hold bonds issued earlier in the year.
9. What the shift means for passive bond investors
Investors with broad, long-duration exposure to investment-grade corporate bonds are the most affected by the trend. Those who buy the market through index funds or similar passive vehicles may be taking a large AI bet without intending to, simply because index weights follow issuance.
As a larger share of AI spending is financed with debt, the health of the corporate bond market becomes more closely linked to whether these infrastructure investments pay off over the coming decades. That is a long time horizon for a technology that is still evolving rapidly.
The issue is less about the immediate creditworthiness of the borrowers, many of which are among the largest and most profitable companies in the world, and more about the structure of portfolios. Concentration risk can build quietly, and it often becomes visible only when the underlying theme falls out of favor.
Investors who want to gauge their own exposure can start by reviewing the maturity profile and issuer breakdown of the bond funds they hold. Funds concentrated in long-dated investment-grade corporate debt are the ones most likely to reflect the shift in issuance that PIMCO has identified.
10. What's next for AI debt and bond portfolios
PIMCO's playbook for investors is straightforward: keep investment-grade quality, but source more of it from agency mortgage-backed securities, higher-rated non-agency mortgage bonds and emerging-market debt rather than relying solely on long-dated corporate bonds increasingly dominated by AI borrowers.
The next signals to watch are the pace of hyperscaler and data center bond sales through the rest of 2026 and how policy developments following the midterm elections affect data center projects. If AI-related borrowers continue to dominate long-dated issuance, the concentration Karoui describes stays a central consideration for anyone holding broad corporate bond exposure.