Some banks that were lining up to fund AI deals a few weeks ago are now telling market participants they would be reluctant to finance additional deals, Reuters reported on 9 November, citing several banking and credit sources. Some lenders have shown that reluctance even when deals are backstopped by names such as Anthropic and Google, the sources said, without naming the lenders.
Anthropic and Google declined to comment, according to Reuters. Two sources told Reuters the change has come in just the past few days.
Business Pill · Financial Conditions
A one-minute explainer of what financial conditions are. It teaches the general idea only and does not discuss the lenders in this piece.
The key insight: Banking and credit sources told Reuters that some lenders are now reluctant to finance additional AI deals, even when they involve backstops from names such as Anthropic and Google. Bank of America’s Yuri Seliger puts the widening in 10-year hyperscaler spreads at about 50 basis points this year against roughly 20 for non-AI industrials, 2.5 times as much, by our arithmetic.
Reuters ties the pullback to the market selloff that erased $2.5 trillion in market value from the Nasdaq 100 on Thursday and to the US midterm results, which it says are adding anxiety about more oversight of data centers.
Chris Poggi, a finance partner at Ropes & Gray, told Reuters: “Over the past year there was unbridled enthusiasm to provide debt to neoclouds, anyone who was building data centers or some sort of AI asset”. He added: “Debt investors’ books have become full and now they are assessing if they have extended themselves as much as they’re going to extend themselves this year.”
Investors and lenders are also stress-testing cash flows for economic downturns and scrutinizing contract terms, including the conditions under which large technology customers could reduce or terminate commitments, two of the sources said, according to Reuters.
Spread widening per Bank of America analyst Yuri Seliger and JPMorgan’s Q3 commercial real estate figures, both as reported by Reuters on 9 November 2026. The 2.5x comparison is our arithmetic.
The Numbers in the Report
The spread on 10-year hyperscaler debt has widened by about 50 basis points this year, against roughly 20 basis points for high-quality industrial companies not tied to AI, according to Bank of America credit analyst Yuri Seliger, as quoted by Reuters. That is 2.5 times the widening, by our arithmetic. Reuters reports that bonds issued earlier this year by Meta, and SpaceX bonds, are trading below par.
At JPMorgan Chase, loans backed by other income-producing properties, a category that includes data centers, stood at $29 billion at the end of the third quarter, Reuters reports. That is 12% of the bank’s commercial real estate exposure, up from 8% three years ago. JPMorgan declined to comment.
Nebius had explored a high-yield bond in September but decided last month not to go ahead, one source familiar with the matter told Reuters. Nebius did not respond to a Reuters request for comment. Reuters notes that Nebius entered a $3.5 billion senior secured loan facility from Citigroup in October.
Supply Is Still Rising
Reuters reports that Goldman Sachs Research projects global hyperscaler debt issuance will reach $420 billion in 2027, more than 60% above its $250 billion estimate for fiscal 2026. In our 8 November piece, MarketWatch put Goldman’s 2027 figure at $440 billion; the two reports give different 2027 numbers, and we report both as published.
Jay Bala, co-founder and CEO of alternative asset manager AIP, told Reuters: “Most banks and credit investors will take a wait and see approach”.
As we read the Reuters report, the caution sits at the margin rather than in the totals: Goldman Sachs Research, as cited by Reuters, still projects hyperscaler debt issuance rising to $420 billion in 2027.
As we read it, the concern Reuters describes is about structure as much as size: lenders are checking the conditions under which large technology customers could reduce or terminate commitments.
Chris Poggi, finance partner at Ropes & Gray, to Reuters, 9 November 2026
“Debt investors’ books have become full and now they are assessing if they have extended themselves as much as they’re going to extend themselves this year.”
Three Implications
FOR BORROWERS Reuters reports that Nebius decided last month not to go ahead with a high-yield bond it had explored in September, and took a $3.5 billion senior secured loan facility from Citigroup in October.
FOR BANKS Reuters reports that JPMorgan’s category including data centers was $29 billion, or 12% of its commercial real estate exposure, at the end of the third quarter, up from 8% three years ago.
FOR BOND INVESTORS Reuters reports that Meta bonds issued earlier this year, and SpaceX bonds, are trading below par.
The framework describes the AI build-out this way: “It is a sovereign-scale financing problem dressed in tech-company clothing”.
As we read it, the Reuters report shows that financing side at work: lenders described by Reuters’ sources are weighing how much more AI exposure their books can take. The framework predates this report and does not discuss these lenders.
What Is Not Established
Reuters does not name the banks that have grown reluctant, and the shift is described by unnamed banking and credit sources. The report does not say any specific Anthropic- or Google-backed deal was refused or repriced.
This Week In The AI Supercycle
The Great Convergence of Capital, Power, and Geopolitics
This week’s issue: for two years the AI build-out was priced as a supply story, and last week as a credit story. This week it was priced as a permission story, and the answer arrived as a cost, not a ban.
Banking and credit sources told Reuters that some lenders are now reluctant to add AI exposure even with Anthropic or Google backstops. Hyperscaler debt spreads have widened about 2.5 times as much as those of non-AI industrials this year, by our arithmetic on Bank of America’s figures.
A note on sourcing. This piece is based on a Reuters report of 9 November 2026 that cites unnamed banking and credit sources; Reuters did not name the lenders described as reluctant. Spread, loan and forecast figures are as reported by Reuters. Comparisons labelled as our arithmetic are FourWeekMBA’s own.
Gennaro Cuofano is a CRO and tech executive who has worked in AI since late 2015, starting with bringing NLP, natural-language generation, voice and chatbot products to the marketing industry. His background is in law and finance: he holds a Master's degree in Law and an International MBA with an emphasis on corporate finance (LUISS Business School and the University of San Diego, 2012), and worked as a financial analyst at a real-estate investment firm in San Diego and as an assistant controller. His work focuses on business model strategy, business engineering and, more broadly, structural analysis: how companies make money, read from their own filings. He created FourWeekMBA and leads research there and at The Business Engineer, his newsletter on AI and business strategy, with over 95,000 subscribers and more than 1,000 published analyses. His writing has also appeared on Entrepreneur (for example, How to Digitize a Family Business in Europe) and HackerNoon. Find him on LinkedIn and Substack. See how we source.
