AI borrowing falls sharply as debt investors reassess risks

By the Editorial Team · Monday, October 12, 2026

AI-related debt financing dropped to $23 billion in September, less than half of August’s total, as investors became wary of the risks from heavy borrowing in the tech sector. This slowdown reflects growing concerns over environmental impacts, regulatory challenges, and rising borrowing costs, prompting investors to reassess their credit exposure and financing strategies.

Global debt issuance linked to artificial intelligence (AI) projects fell sharply in September as investors reassessed the risks associated with heavy borrowing across the technology sector. According to data compiled by Morgan Stanley, AI-related debt financing totaled $23 billion last month, less than half the amount raised in August. This marks a significant slowdown following a peak in June when companies secured a record $113 billion to fund infrastructure such as data centers and specialized chips crucial for AI development.

The decline was particularly notable in the United States investment-grade market, where blue-chip technology companies had borrowed approximately $306 billion between January and August. In that segment, AI-related bond issuance effectively stalled in September.

Morgan Stanley attributed the downturn mainly to the considerable volume of debt already accumulated earlier this year. Investor concerns have increased regarding the return on such heavy investments, especially amid growing political opposition to data center construction due to environmental and community impact issues. Energy consumption, water supply, and air quality have emerged as key points of resistance against new data center projects, particularly in the US.

One high-profile example is Oracle’s data center campus in New Mexico, which has faced difficulties securing access to electricity, putting approximately $18 billion in associated loans under pressure.

Industry experts suggest the reduced pace of financing may allow investors to reexamine their credit exposure. John Aylward, founder of London-based credit firm Sona Asset Management, commented at a recent private capital summit that ongoing financing has become more challenging as lenders tighten their standards. Jordan Rieger, managing principal at Monarch Alternative Capital, described the slowdown as a "healthy breathing opportunity" for debt investors to reassess the terms and structures of existing financing arrangements.

Despite the cautious stance, some sizable AI-related financing deals remain in the pipeline. Wall Street banks are reportedly preparing a $60 billion package to support Broadcom’s acquisition of Anthropic, while SpaceX is in negotiations to raise $40 billion to purchase Nvidia chips.

Overall, AI-related debt issuance has surged to roughly $466 billion so far in 2026, up sharply from $101 billion during the same period last year. This increased supply of debt has also led credit investors to demand higher yields, pushing up borrowing costs in the sector. The recent slowdown illustrates a market recalibration as investors weigh the risks and returns of continued heavy investment in AI infrastructure amid economic and regulatory uncertainties.

This article was reported by synthesizing coverage from multiple independent news outlets to present a balanced, unbiased account. Read about our editorial standards. Spotted an error? Let us know.