Amazon lines up $17.5B Citibank loan as AI infrastructure spending accelerates
Amazon has arranged a massive new credit lifeline, securing a $17.5 billion delayed draw term loan facility led by Citibank and a syndicate of other major banks. The financing, disclosed in a June 10 filing with the U.S. Securities and Exchange Commission, significantly expands the company’s borrowing capacity at a time when its capital spending and artificial intelligence ambitions are surging.
The loan is structured as a senior unsecured delayed draw term loan, meaning Amazon is not required to take the cash all at once. Instead, it can tap the funds in tranches as needed up to a preset deadline. According to the filing, bank commitments under the facility remain available until September 30. Amazon can draw down any portion of the $17.5 billion before then, unless it decides to access the full amount earlier.
Once Amazon borrows from the facility, each draw will mature three years from the date the funds are taken. Citibank N.A. is acting as the administrative agent on the agreement, coordinating the banking group and managing the mechanics of the credit line.
Officially, Amazon framed the loan in broad terms. In the filing, the company said simply that “borrowings under the DDTL Facility will be used for general corporate purposes.” That language gives management maximum flexibility: the money is not earmarked for a single project or acquisition but can be reallocated as priorities shift, whether for investment, operations, or balance sheet needs.
Privately, however, Amazon has offered more color on how it might deploy the funds. As reported by Bloomberg, a company spokesperson said the proceeds could be used to support new investments, cover capital expenditures, and refinance or repay existing debt. In practice, that mix of uses aligns closely with Amazon’s current strategic focus on artificial intelligence, cloud infrastructure, and high-performance computing.
The timing of the facility closely tracks Amazon’s huge ramp-up in AI-related spending. The company has telegraphed plans to pour roughly $200 billion this year into AI infrastructure and other capital projects, spanning data centers, advanced chips, networking gear, and the broader cloud backbone that underpins its AWS business. That capital-intensive buildout is designed to support both Amazon’s own AI products and services and those of its enterprise customers.
Part of that strategy involves backing leading AI developers. The report connected the new loan facility to Amazon’s broader technology infrastructure and AI investment roadmap. It noted that Amazon has already put about $10 billion into Anthropic, a prominent AI startup, with an additional $15 billion commitment potentially on the table. These investments are meant to ensure that sophisticated AI models are trained and deployed on AWS, reinforcing Amazon’s position as a core platform for AI workloads.
The same report also highlighted expectations that Amazon could commit tens of billions of dollars overall into AI-focused partnerships and companies over the coming years. While specific figures and counterparties can shift, the message to markets is clear: Amazon sees AI as a foundational pillar of its future growth and is willing to leverage its balance sheet to stay competitive in what has become a capital-heavy arms race among major tech players.
Beyond equity stakes and partnerships, Amazon continues to plow money into traditional infrastructure: more cloud regions, larger data centers, and expanded computing capacity. These physical and digital assets are essential to hosting generative AI models, running inference at scale, and offering customers the low-latency, high-availability services they now expect. A large, flexible credit line gives Amazon an extra cushion to execute these multiyear buildout plans without being forced into unfavorable financing conditions.
The new loan also fits into a broader pattern of Amazon actively tapping global debt markets. Just days before the SEC filing, on June 8, the company sold 14 billion Canadian dollars of high-grade bonds, an issuance worth roughly $10 billion at the time. Since March, Amazon has placed bonds in multiple currencies, including euros, U.S. dollars, and Swiss francs, signaling a strategy of diversifying its funding sources and investor base.
The company has not indicated that the $17.5 billion facility is meant to replace or refinance any particular bond offering. Nor has it disclosed that it has drawn on the loan so far. For now, the credit agreement appears to function as an additional layer of financial flexibility, sitting alongside unsecured bond issuance and existing cash on the balance sheet.
From a legal and structural standpoint, Amazon described the DDTL Credit Agreement as containing the usual suite of representations, warranties, covenants, and default provisions typical for a large corporate borrower. Notably, the filing stated that the deal does not include specific financial covenants, which means Amazon is not bound to maintain particular leverage or coverage ratios to keep the facility in place. That lack of financial maintenance tests is common for highly rated, cash-generative issuers.
If an event of default does occur, Amazon would generally receive specified grace periods to remedy the situation. Should a default remain uncured or not be waived by lenders, the outstanding principal could become immediately due and payable. In that scenario, lenders would also have the right to terminate their commitments to provide any undrawn amounts. These protections are standard features, especially in facilities led by large global banks.
In the filing, Amazon also characterized the participating banks as full-service financial institutions providing a range of activities beyond lending. Their business lines can include trading, commercial and investment banking, advisory services, asset management, research, hedging, brokerage, and market-making. This highlights another strategic layer: deep banking relationships do not just unlock credit, but also underpin capital markets access, M&A advisory, and risk management services that become increasingly important as a company scales.
This loan arrives at a key moment in the corporate funding environment. AI infrastructure buildouts require extraordinary upfront cash, often long before the full revenue impact is realized. At the same time, higher interest rates and tighter financial conditions have made capital more expensive and selective. By securing a large, multi-year facility now, Amazon mitigates some of that macro risk, ensuring it can keep executing on its roadmap even if markets become more volatile.
There is also a competitive dimension: major peers in the tech sector are racing to line up their own war chests for AI. From chip manufacturing to data center construction to long-term commitments with AI research labs, the scale of planned spending across the industry runs into hundreds of billions of dollars. Amazon’s decision to bolster its liquidity through both bond issuance and bank lending signals to investors and rivals that it intends to remain a major player in that contest.
For creditors, the structure of a delayed draw term loan is attractive because it allows banks to commit capital with a clear maturity and defined amortization profile, while still giving the borrower flexibility on timing. For Amazon, the key advantage is optionality: it only starts paying interest on amounts it actually draws, and it can align those drawdowns with specific project timelines or refinancing windows.
The facility could also play a role in smoothing Amazon’s debt maturity schedule. As bonds issued in prior years come due, the company can use drawn amounts to retire older, potentially more expensive obligations, then refinance opportunistically in the public markets when conditions are favorable. That mix of bank and bond financing helps maintain a balanced capital structure and preserves room for future issuance.
From an investor’s perspective, the announcement underscores just how capital-intensive Amazon’s next phase of growth will be. The e-commerce and cloud giant is no longer simply scaling logistics networks and server farms; it is now building AI supercomputing infrastructure, specialized chips, and global data center clusters optimized for machine learning workloads. All of that requires not just technology vision, but careful treasury management.
While the SEC filing itself avoided naming specific AI projects or products, the broader context makes the link clear: AI is now deeply embedded in Amazon’s corporate strategy, and financing decisions are increasingly being made with that lens in mind. Whether through strategic stakes in AI startups, expanded AWS offerings, or internal tools that streamline retail, logistics, and advertising, the company is positioning itself for an AI-driven future-and is willing to commit substantial financial firepower to do so.
Taken together, the $17.5 billion Citibank-led loan, the multi-currency bond issuances, and the vast AI infrastructure budget paint a picture of a company preparing for the next decade, not just the next quarter. By locking in large-scale, flexible financing now, Amazon is giving itself room to pursue aggressive AI and cloud expansion while maintaining the resilience and optionality that have long been hallmarks of its financial strategy.