Barclays explores insurance protection for more than $10 billion in AI infrastructure debt
Key Facts
Bloomberg reported on October 28 that Barclays is sounding out insurers about buying protection against default on more than $10 billion in debt earmarked to finance AI infrastructure. The potential move aims to make room for more lending as hyperscalers’ financing needs grow. Bloomberg said discussions remain exploratory and the bank has made no final decisions on the timing or size of any potential deal. That puts the initiative at the stage of exploring risk-transfer options, rather than a transaction that has already been completed. Bloomberg reported that the stated aim is "to create capacity for more lending as hyperscalers’ financing needs escalate."
The scale of the debt being considered for coverage reflects the expansion of financing tied to data centers and AI. Barclays Private Bank said in an October market commentary that borrowing by the largest hyperscalers and AI-related entities had risen from slightly more than $50 billion in 2022 to nearly $300 billion so far in 2026. That growth provides context for banks’ search for ways to sustain financing while distributing credit risk. The International Monetary Fund estimated that these companies’ AI-related capital spending could reach $3.4 trillion through 2029, potentially putting pressure on their balance sheets in the coming years. The fund also said hyperscalers had raised more than $100 billion in bond financing since January 2025, alongside leveraged loans and intercompany financing arrangements.
Infrastructure financing channels are varied. The IMF pointed to potential demand for private credit, with insurers and asset-based financing playing a role in infrastructure debt. This context puts insurers among the entities that could help finance the sector or absorb some of its risks, as Barclays gauges their interest in default protection on the debt. But the fund also pointed to execution hurdles to expansion, including data centers’ grid connection waits, which can range from 2 to 7 years. Those waits heighten the importance of financing large projects that require coordination across infrastructure, energy and finance. In the securitization market, data center-backed issuance has totaled $46 billion since 2018, and the fund forecast it would reach $150 billion by 2028.
Credit indicators show a sharp divergence in how risk is priced across the AI sector. On October 15, Oracle’s 5-year credit default swap spread widened by 26 basis points overnight to a record 288 basis points, according to Reuters. On the same day, the equivalent spread for data center operator CoreWeave topped 1000 basis points, Reuters reported, signaling investors’ higher assessment of the company’s credit risk. Reuters Breakingviews reported in September that data center debt issued during the year had reached nearly $500 billion by early August, while lenders were demanding higher pricing and stronger protection. These developments provide context for lenders’ efforts to manage exposure to rising financing, without by themselves establishing the terms or cost of any protection Barclays is considering.
As for the stock, EL7 data showed Barclays shares, listed under the ticker BCS, closed at $22.46 on October 27, compared with a close of $22.27 on October 26 and $21.74 on October 23. These figures are daily closing prices in dollars and do not represent intraday movement or a price during the October 28 session. Nor do they alone establish the impact of exploratory discussions on the stock’s trading. In the broader context of financing risks, the IMF reviewed 14 disclosed circular financing deals since 2023 and found about $300 billion in equity investments and $650 billion in debt against $1.4 trillion in future revenue commitments. The next step for Barclays will depend on how discussions develop, as the bank has not announced a final decision on the timing or size of the potential deal.