The four largest US technology companies have placed an unprecedented bet on artificial intelligence, locking in nearly $2.4 trillion in combined purchase commitments, contractual obligations, and long-term leases, according to a recent report.
The scale of AI commitments
Alphabet, Amazon, Meta, and Microsoft are at the center of an AI infrastructure build-out that dwarfs previous technology investment cycles. Alphabet leads the group with roughly $900 billion in total commitments, a figure about nine times higher than a year earlier. Meta follows with around $700 billion in future spending. Amazon has budgeted $220 billion in capital expenditure for this year alone, and Microsoft's obligations round out the total.
These figures cover far more than ordinary annual capital expenditure. They include multi-year purchase orders for advanced chips, long-term power purchase agreements, and data center leases that stretch decades into the future. Much of this spending sits in financial footnotes for now, but it will eventually move onto balance sheets as the obligations mature.
Meta's long-term lease exposure
Meta's future spending is particularly concentrated in data center leases. About half of the company's total commitments are tied to leases with terms as long as 30 years. That is roughly eight times the lease obligations Meta carried a year ago. The company has been aggressively expanding its infrastructure to support AI workloads across its social platforms and cloud ambitions.
Alphabet's commitments have also grown rapidly, with a significant portion structured through off-balance-sheet vehicles. Those structures keep headline debt figures looking manageable while allowing the company to secure the enormous physical and energy resources needed for AI data centers.
Cash flow pressure emerges
Two of the four companies are already showing signs of strain. Alphabet reported negative free cash flow of nearly $6 billion in the second quarter, its first quarterly outflow in close to two decades. Amazon's free cash flow turned negative on a trailing-twelve-month basis earlier this year, and its second-quarter results showed the cash position remained under pressure even as AWS revenue grew 37 percent.
Meta is expected to follow a similar path. The company narrowed its 2026 capital expenditure forecast to between $130 billion and $145 billion by raising the low end of the range. Its free cash flow fell 91 percent year over year in the second quarter. The pattern across all four companies is consistent: revenue is growing, but capital spending is growing faster, and the gap shows up first in cash flow.
Cloud growth remains strong
Despite the pressure, demand signals remain positive. AWS reported its fastest growth in four years this week. Alphabet's cloud unit grew 82 percent in the second quarter. These results suggest that enterprise customers are still expanding their AI usage, though the question is whether the pace of revenue growth can eventually match the pace of infrastructure spending.
The top cloud operators are collectively on track to see capital expenditures overtake the cash their core businesses generate. That threshold has never been crossed before, and it represents a fundamental shift in how these companies operate.
The AWS analogy
Amazon CEO Andy Jassy has compared the current moment to the early days of AWS. During the first AWS build-out, the company spent years investing heavily before demand caught up. That bet eventually paid off, with AWS growing from a side project into a business that generates more than $40 billion per quarter. Jassy's analogy is meant to reassure investors that the current AI spending will follow a similar trajectory.
The difference is scale. The original AWS build-out cost a fraction of what the industry is now committing in a single year. The current AI infrastructure cycle involves not only computing hardware but also energy infrastructure, cooling systems, and long-term real estate. The complexity and capital intensity are orders of magnitude larger.
How accounting smooths the pain
All four companies remain profitable for now. That is partly because capital spending is depreciated over years rather than booked as an immediate expense. Accounting rules allow companies to spread the cost of data centers, chips, and other infrastructure across their useful lives. This smooths the impact on reported earnings, but it does not eliminate the eventual cost.
Depreciation charges will rise in future quarters as more infrastructure comes online. That will pressure operating margins even if revenue growth continues at a healthy clip. Investors are watching these numbers closely, because the gap between capital commitments and operating cash flow is a leading indicator of future financial stress.
What the commitments include
The $2.4 trillion figure is not a single line item. It includes:
- Long-term purchase orders for AI accelerators and other semiconductors
- Power purchase agreements with utilities and independent power producers
- Data center leases ranging from several years to three decades
- Construction and engineering contracts for new facilities
- Network infrastructure and interconnection agreements
These commitments are legally binding. Whether AI revenue materializes at the scale these companies are betting on or not, the money will come due. That is the core risk embedded in the current cycle.
Historical context
The current spending wave has no direct precedent. During the dot-com boom, telecommunications companies accumulated large debts to build fiber networks, only to see many of those investments become stranded when demand did not materialize as expected. The aftermath included bankruptcies, consolidation, and a long period of reduced capital spending.
In the cloud computing era, the major providers built data centers more incrementally. AWS, Microsoft Azure, and Google Cloud expanded capacity in response to visible demand, and the capital intensity of the business was manageable relative to cash generation.
The AI cycle is different in both scale and speed. Companies are building capacity well ahead of proven demand, and they are doing so collectively rather than individually. This creates a risk of overcapacity if AI adoption slows or if technological breakthroughs reduce the need for giant clusters.
The market reaction
Investors have so far rewarded the companies for their AI commitments. Share prices for all four have risen significantly over the past two years. The market appears to trust that AI spending will create durable competitive advantages and that the companies will eventually generate strong returns on their infrastructure investments.
But the cash flow deterioration is starting to get attention. Analysts have begun to ask whether the scale of commitments is justified by near-term revenue opportunities. Some have questioned the use of off-balance-sheet vehicles, which make it harder for investors to see the full extent of liabilities.
The road ahead
All four companies remain committed to their AI strategies. Alphabet continues to invest across its cloud and consumer AI products. Amazon is expanding AWS capacity and integrating AI into its logistics and retail operations. Meta is building out compute infrastructure for content recommendation and generative AI features. Microsoft is leveraging its partnership with OpenAI and its Azure platform to secure enterprise workloads.
The next several quarters will be critical. Revenue growth will need to accelerate, or capital spending will need to moderate. Either path carries consequences. Accelerating growth requires continued economic expansion and customer adoption. Moderating spending risks losing ground in a competitive race where infrastructure leadership is seen as essential.
The $2.4 trillion in commitments represents a collective judgment that AI will reshape the global economy. That judgment may prove correct, but the financial journey will be a test of balance sheet strength, operational discipline, and investor patience. The companies have placed their bets. Now they must deliver.
Source: TNW | Amazon News