You guys — this just hit TechCrunch and it's the kind of scale that should stop your scrolling for a minute because Amazon literally just raised $17.8 billion in new senior notes through bookrunners Morgan Stanley and Citi! Here is how the math actually breaks down: they structured $24 million of those as fixed-rate debt at a 6.7% coupon, while the remaining $15.8 billion came out as floating rate (SOFR plus margin) with maturities slated for late December 2032 and early March 2033 respectively. That's not some small tech budget — that's an enterprise-scale capital raising exercise on a level most of us don't even think about in our day-to-day lives, and it was driven almost entirely by the need to continue pouring cash into AI infrastructure investments. When you see this volume being deployed for one initiative, it tells you everything you need to know about where their bets are placed right now.
But wait — there's more debt on top of that already sitting in the vault like a loaded gun! Amazon also maintains an existing $25 billion undrawn credit facility with BMO and JPM Chase (which by the way, is still largely untapped) AND they just added another new revolving line from Goldman Sachs. That means their total senior note debt now sits around $194 billion before you even factor in other lines of credit, bringing their overall borrowing capacity to roughly $219 billion. This isn't a company that has "too much" cash sitting on the sidelines; this is an organization aggressively leveraging itself because they believe the AI CapEx engine will generate outsized returns. They are betting the farm — literally beting nearly two hundred billion dollars - that generative models will become central to their business model and deliver massive ROI, which isn't a casual commitment by any sane CEO of any company.
What I take from all this is pretty profound: Amazon has moved past "AI as an experimental feature" and fully integrated it into core operational strategy at the highest levels of financial planning. The sheer amount of debt they just issued specifically for AI infrastructure investment means they aren't guessing whether LLMs will be useful; they are operating with near-certain confidence that generative models are a generational technology worth anchoring their entire future on. This is what I mean about capital following conviction — when the biggest company in retail and cloud decided this was where to spend, the financial markets responded by letting them issue bonds at 6.7% for twenty billion bucks of it. If you're still wondering whether the AI hype train has any legs left, look at how hard Amazon is pushing on its throttle. The betting house isn't close — it's about as large and permanent as the company itself now.
Source: https://techcrunch.com/2026/06/10/fresh-off-bond-sale-amazon-borrows-17-5-billion-from-banks-as-ai-spending-continues
But wait — there's more debt on top of that already sitting in the vault like a loaded gun! Amazon also maintains an existing $25 billion undrawn credit facility with BMO and JPM Chase (which by the way, is still largely untapped) AND they just added another new revolving line from Goldman Sachs. That means their total senior note debt now sits around $194 billion before you even factor in other lines of credit, bringing their overall borrowing capacity to roughly $219 billion. This isn't a company that has "too much" cash sitting on the sidelines; this is an organization aggressively leveraging itself because they believe the AI CapEx engine will generate outsized returns. They are betting the farm — literally beting nearly two hundred billion dollars - that generative models will become central to their business model and deliver massive ROI, which isn't a casual commitment by any sane CEO of any company.
What I take from all this is pretty profound: Amazon has moved past "AI as an experimental feature" and fully integrated it into core operational strategy at the highest levels of financial planning. The sheer amount of debt they just issued specifically for AI infrastructure investment means they aren't guessing whether LLMs will be useful; they are operating with near-certain confidence that generative models are a generational technology worth anchoring their entire future on. This is what I mean about capital following conviction — when the biggest company in retail and cloud decided this was where to spend, the financial markets responded by letting them issue bonds at 6.7% for twenty billion bucks of it. If you're still wondering whether the AI hype train has any legs left, look at how hard Amazon is pushing on its throttle. The betting house isn't close — it's about as large and permanent as the company itself now.
Source: https://techcrunch.com/2026/06/10/fresh-off-bond-sale-amazon-borrows-17-5-billion-from-banks-as-ai-spending-continues