Moody’s Ratings, a leading credit ratings and financial research provider, has reported that US life insurers are increasing their financial exposure to artificial intelligence (AI) infrastructure, although the investments account for a relatively small proportion of their overall assets.
The agency, in its latest study noted that growing investment in data centres, computing equipment and major technology companies could create greater concentration risks for some insurers as the sector expands.
A news report from Reinsurance News, a risk-underwriting-focused online medium, quoted Moody’s as saying that the US life insurers hold approximately USD $85.2 billion in investments linked to the AI ecosystem at the end of the second quarter of 2026, representing about1.4% of their cash and invested assets, marking a 24% increase from the USD $69 billion recorded at the end of 2025.
Moody’s analysis examined three main areas of exposure, namely debt issued by major technology companies, financing for digital infrastructure projects, and investments associated with semiconductor manufacturers.
- Advertisement -
The agency identified Alphabet, Amazon, Meta Platforms, Microsoft and Oracle as the five major US technology companies central to its assessment of the AI infrastructure expansion.
The Moody’s report’s findings suggested that the immediate effect of these investments on insurers’ portfolios was likely to remain limited. However, the agency expects exposure to increase as demand for computing capacity drives further spending on AI infrastructure and also estimates that global investment in the AI ecosystem could require at least USD $3 trillion over the next five years, with broader estimates ranging from USD $3 trillion to USD $7 trillion.
Moody’s also projected that US life insurers could increase their AI-related holdings by between USD $50 billion and USD $100 billion by the end of 2027, assuming global digital infrastructure spending reached USD $1 trillion during that year, projecting that insurers are likely to remain an important source of funding for the investment-grade debt used to finance these developments.
The study reflected that AI-related investments accounted for 2.1% of US life insurers’ total fixed-income holdings in the second quarter of 2026, up from 1.7% at the end of the previous year; and Moody’s expects this share to rise further during the remainder of 2026 and into 2027, reflecting continued investment in computing infrastructure and related assets.
The agency further reported that debt issued by the five major technology companies represented approximately USD $32.2 billion of insurers’ AI-related investments at the end of the second quarter, representing an increase of 21% compared with the end of 2025 and equated to around 6.6% of the companies’ outstanding debt.
Moody’s revealed that Oracle accounted for approximately USD $10 billion of these holdings, while Amazon represented a further USD $8.5 billion, adding that insurers’ purchases during the first half of 2026 has increasingly favoured higher-rated companies, including Alphabet, Amazon and Meta Platforms.
But then, it noted that direct debt holdings did not provide a complete picture of insurers’ exposure. Other investments may be linked to major technology companies through guarantees, leases and credit support arrangements, adding that these connections could leave insurers exposed to the financial performance of individual companies even where the debt is not held directly.
The agency also highlighted the potential for greater concentration within investment-grade corporate debt portfolios as major technology companies issue more debt to finance their expansion. Although these companies generally have strong credit ratings, saying that some insurers hold larger positions than others, making individual exposure levels an important consideration.
The Moody’s analysis revealed that financing for data centres and associated computing equipment was the fastest-growing category and that US life insurers held approximately USD $39.5 billion in digital infrastructure debt at the end of the second quarter of 2026, representing a 34% increase from the end of 2025.
The agency listed these investments as including funding for data centre buildings, electricity infrastructure, graphics processing units (GPUs) and tensor processing units (TPUs), alongside asset-backed securities, commercial mortgage-backed securities and other financing arrangements. Moody’s said most of the debt remained investment grade, although the proportion rated below investment grade increased from 4.7% at the end of 2025 to 6.1% in the second quarter of 2026.
According to Moody’s, the repayment of much of this debt ultimately depends on a relatively small number of companies operating across the AI sector. These include major technology groups, specialist cloud computing providers such as CoreWeave, AI developers including OpenAI and Anthropic, and semiconductor manufacturers such as NVIDIA, Broadcom and Advanced Micro Devices.
The agency cautioned that the range of financing structures used to support infrastructure development could make insurers’ total exposure difficult to quantify as some investments are held through special purpose vehicles, joint ventures, alternative investment funds, commercial mortgage loans and collateralised loan obligations.
Consequently, the agency reported that its figures were likely to underestimate the US life insurance industry’s overall exposure to the AI ecosystem.
Moody’s also identified future spending commitments as another area for monitoring. as the five major technology companies had disclosed approximately USD $2.8 trillion in future commitments, including leases yet to commence, purchase agreements and guarantees. This compared with USD $1.2 trillion in 2025, an increase of around 130%.
Noting that although some of these commitments are not recognised as current liabilities under US generally accepted accounting principles, Moody’s projected that they could result in substantial future funding requirements. The agency added that companies would probably need to raise additional debt where planned capital expenditure exceeded the cash generated by their operations.
It maintained that the expansion of AI infrastructure also faced practical and regulatory challenges. Moody’s pointed to constraints on electricity supply, planning and permitting requirements, and growing opposition to data centre developments in some US communities, noting that these factors could delay construction, increase financing costs and affect the cash flows available to repay project debt.
The Moody’s study report referred to regulatory developments in New Jersey and Texas during 2026, alongside delays affecting individual infrastructure projects, adding that increasing demand for electricity is placing pressure on some local power networks, while governments were assessing the effects of data centre construction on energy use, water resources and the environment.
In addition, another concern identified by the agency was the interconnected nature of financing across the AI industry. Companies may invest in one another, purchase each other’s products or provide financing linked to future commercial relationships. For example, it reported that a special purpose vehicle could borrow money to acquire computing equipment and then lease it to a technology company, with the anticipated rental income supporting repayment of the debt.
Moody’s clarified that these arrangements could create interconnected counterparty risks, meaning financial difficulties affecting one transaction could have consequences for other investments; and that semiconductor manufacturers that provide financing to customers purchasing their products were another example of the relationships highlighted in the report.
The agency reported that insurers would need to monitor the credit quality of major technology companies, the potential effects of regulatory restrictions and local opposition to data centre developments, and the risks arising from interconnected financing arrangements.
In its concluding remarks, Moody’s stressed that AI-related investments remained a relatively modest component of US life insurers’ overall assets, but their rapid growth warranted continued attention.
It projected that as investment in digital infrastructure continued to increase, the distribution of exposure between borrowers, the structures used to finance projects and the financial links between companies could become increasingly important considerations for insurers.
