Bonds, including whoppers such as Alphabet’s $32bn offering in February, remain a popular choice. Even for these everyday instruments, however, the scale required means issuers must get creative. Foreign credit markets are increasingly sought out. So far this year about 30% of bonds issued by the five biggest cloud providers—meaning the hyperscaler quartet plus Oracle— have been in non-dollar currencies, including euros, pounds and yen. At the same time, other instruments are growing in popularity. SPVs are now commonplace. One example is Sopaipilla, a project masterminded by Meta (and named after a pastry). It consists of ten data centres in El Paso, Texas, which in total will provide roughly a gigawatt of computing power. The site is owned by a company called Wurldwide, which in turn is owned by Meta, with a 20% stake, and BlackRock, an asset-management giant, which holds the remaining 80%. Wurldwide issued $13bn of debt in August to fund the building of the campus, which Meta has agreed to lease for 20 years. Such structures tick plenty of boxes. BlackRock and its clients get a stream of steady income guaranteed by Meta, which in turn gets to add computing capacity while keeping the debt off its balance-sheet. Credit-rating agencies have given the SPV a score one notch below Meta’s. Chip-backed financing, which entails using the silicon deployed in a data centre as collateral, is also on the rise. SpaceX is said to be looking to secure $40bn of funding using the approach in order to purchase more chips from Nvidia. Often chip-backed financing is arranged through an SPV. Oracle is in talks with Apollo, a private-markets giant, and Goldman Sachs, an investment bank, to fund a big purchase of semiconductors that will be held by a separate entity and leased to Oracle. Anthropic and OpenAI, two makers of frontier AI models who are eager to gain greater control over their computing capacity, are reportedly negotiating similar arrangements to procure large amounts of chips directly. Chip-backed financing makes assessing risk “infinitely more complicated”, says Roelof Steenkamp of Fitch, a credit-rating agency. A consequence of such arrangements is that a data-centre’s bricks and mortar, the chips inside it and even the site’s power source (if it has its own) can each be financed by its own layer of debt.

On top of this are the obligations that are not quite debt, but close to it. The five biggest cloud providers have lease liabilities on their balance-sheets totalling over $250bn, according to CreditSights, a research firm. A further $1.2trn-worth of leases have been agreed but are yet to commence. Then there are commitments to purchase products from suppliers worth an additional $1.5trn. A final complexity comes from the backstops that companies in the AI nexus with bigger and healthier balance-sheets are offering to guarantee the debts of others. Nvidia has been particularly enthusiastic in this regard, placing itself on the line to cover as much as $300bn of debt used to purchase its chips if its customers are unable to repay their creditors. Alphabet and Broadcom, which develops custom chips for cloud providers and model-makers, have also provided backstops to some of their customers. Such deals make it even more difficult for investors to determine where the risks lie, and how far losses from a default might spread. Who is exposed? The short answer is everyone. All types of financial firms are holding AI debt, partly because there is so much of it. Large asset managers have been the biggest buyers. For example, Pimco, one of the world’s biggest bond-fund managers, bought $18bn-worth of the $27bn bond issued for Meta’s Hyperion data-centre campus in Louisiana. It then sold on tranches of the debt to institutional investors, such as pension funds and insurance firms. Providers of private credit, including Apollo and Blackstone, are spearheading many of the more creative deals. Sometimes these loans, too, are syndicated to institutional investors—including insurers controlled by the private-credit providers themselves. A sizeable chunk of the debt Apollo lent to Broadcom in June reportedly ended up on the books of Athena, its insurance arm. Hedge funds have also piled in to the market for AI debt, showing particular interest in the neoclouds. Despite the recent market wobble, investors on the whole are not yet panicking about the debt binge. Only about 15% of the value of AI-related bonds issued this year have junk status. And although the various backstops in place make the stock of AI debt more difficult to interrogate, the cash-rich balance-sheets of companies like Nvidia would help to absorb any shock. On

October 15th Jensen Huang, Nvidia’s boss, defended the prospects of CoreWeave, some of whose debts his company has guaranteed, arguing that its “cost of capital is actually decreasing” over time. Even so, dangers lurk ahead. One threat is from rising borrowing costs. War in the Gulf, persistent inflation and anxiety over the growth of public debt have caused yields on government bonds to climb rapidly in recent months, pushing up interest rates for corporate borrowers. The scale of debt issued in support of the AI boom has itself played a part. For the deep-pocketed hyperscalers, that is a source of pain; for shakier borrowers like CoreWeave —whose interest bill already soaks up a quarter of annual revenue—it could prove deadly. The even bigger threat would be if demand for AI were to stagnate. Data from Ramp, a provider of corporate charge cards, shows that spending on AI among its customers has flatlined since July, as businesses try to contain spiralling costs. Softening growth, in turn, might lead Anthropic or OpenAI to delay indefinitely their much-anticipated public listings, which could cause investors to rethink their exposure to AI. The bigger the tower of AI debt grows, the more vulnerable it will become—and the graver the consequences should it come crashing down. ■ Correction (October 22nd): An earlier version of this article understated the size of listed neoclouds’ debts relative to their revenues. Sorry. This article was downloaded by zlibrary from https://www.economist.com/business/2026/10/18/the-ai-industrys-gigantic-debts-are- making-investors-nervous

Business · Business | Call my agent

Personal AI assistants could soon reshape the digital economy But they will first have to prove they are safe for users and helpful for businesses Oct 22nd 2026 Silicon Valley has long dreamed of building a digital assistant capable enough to rid users of many of the tedious personal tasks that gobble up time and attention. Fictional representations, such as J.A.R.V.I.S. (Just a Rather Very Intelligent System) in the “Iron Man” films, set expectations high. In 2010 Apple, the iPhone-maker, acquired Siri, one such aide. It turned out to be disappointingly moronic. Now, amid the current wave of investment in artificial intelligence, the era of personal AI agents may at last have arrived. In the past two months Meta, a social-media giant, and Instinct, a startup, have released two such bots to great fanfare. Instinct’s assistants, which users contact through Apple’s

iMessage service or Meta’s WhatsApp, are already checking users into flights, sourcing sold-out concert tickets and arranging dinners with friends, among much else. Muse, Meta’s offering, has similar powers, but can be reached via WhatsApp, a web browser, a mobile app—or, from later this year, Meta’s smart glasses. The big shift from earlier generations of digital assistants, explains M.G. Siegler, an investor and blogger, is that these new bots can work proactively, rather than reacting only to what users explicitly ask for. Instinct “really just works in the way I believe we all want AI to act”, says Noah Shinn, the startup’s 24-year-old founder. Uptake is soaring, so much so that Mr Shinn reckons “in a few years every user in the world that has an internet connection is going to have access to a product like Instinct.” Before then, however, the companies developing personal agents will have to overcome a variety of hurdles to more widespread adoption, including concerns over security and reliability and the reluctance of some big businesses to allow the tools to interact with their systems. Instinct and Meta have pursued very different approaches to attracting users, but both have been wildly successful. In Instinct’s case, users must be invited by a friend or hand-picked from a waitlist. Even so, the number of users has been growing by about 10% daily. Simplicity keeps them coming back: around half of people who have signed up to Instinct continue using it at least once a week after the first month, according to the startup. Giving Instinct’s agents an instruction is as simple as writing a short text; they will then reply with updates on their progress. If Instinct is the private party taking place behind a velvet rope, Muse is the open-invite rave anyone can partake in (so long as they live in America or Canada). By all accounts, the party is heaving—and people are lingering, too. More than 10m people have downloaded the Muse app since its debut a little over a month ago, according to Sensor Tower, a data provider. ChatGPT, the hit chatbot from OpenAI that precipitated the AI boom, had only 3.7m downloads of its mobile app in the same timeframe after its release, around six months after the web version was first unveiled. Muse is retaining users at about the same rates as ChatGPT, too.