Microsoft, Meta, Oracle, Amazon, and Alphabet have collectively committed to approximately $1.09 trillion in future lease obligations, primarily for AI data centers, according to a report by Reuters based on the companies' financial disclosures.

These lease agreements are not yet active, meaning the amounts are not currently reflected on the balance sheets. Journalists report that the recognized obligations of these five companies stand at about $285 billion.

It is important to note that the $1.09 trillion figure cannot be directly added to existing debt, as these are undiscounted payments spread over several years. However, it does indicate the scale of AI infrastructure that has already been contracted ahead of actual operational start dates.

Identifying Risks

Microsoft has revealed the largest portfolio of future lease payments at $329.1 billion, compared to $88.52 billion in already recognized obligations. Meta reported $278.99 billion and signed an additional $68 billion in data center leases in July. With these new agreements, the total known portfolio for the five companies has risen to approximately $1.16 trillion.

Alphabet disclosed $85.2 billion in future lease obligations, while Amazon reported $137.21 billion. Reuters highlighted Oracle as particularly vulnerable, with $260 billion in future obligations, nearly seven times its already recognized $37.89 billion. Most of Oracle's commitments are related to data centers expected to become operational in the fiscal years 2027 to 2029.

As of the end of May, Oracle's debt was estimated at about 4.4 times its EBITDA. When factoring in existing obligations, this ratio increased to approximately 5.7 times EBITDA.

S&P Global Ratings downgraded Oracle's credit rating to BBB- in July, just above speculative grade. Reuters noted that the agency considered Oracle's $260 billion in future lease obligations in its adjusted debt forecast, anticipating a leverage ratio of around 4.4 in the fiscal year 2027.

Another risk involves reliance on a single major client. According to Reuters, nearly half of Oracle's remaining contractual obligations are tied to OpenAI. Should demand from this client decrease or if contract terms do not align with the lease timelines for data centers, Oracle may face increased financial pressure.

Connection to the AI Bubble

The rise in obligations has sparked discussions about whether the AI boom could lead to credit risks. Companies are constructing data centers in anticipation of demand for computing power, but many of these obligations will not take effect for several years. If computing demand continues to rise, these facilities could become the backbone of the cloud business; however, if demand falters, firms may face long-term payments for unused infrastructure.

Arthur Hayes, co-founder of BitMEX, compared the AI infrastructure boom to the credit cycle of 2008 in an essay. He believes investors mistakenly view data center construction as a high-multiplier technology business, despite a significant portion of expenses resembling real estate debt financing.

Hayes suggests that a slowdown in AI capital expenditure growth could reveal which companies struggle with their debt burdens, potentially prompting authorities to support the sector with liquidity. He predicts that such a reaction might eventually drive Bitcoin to $1 million.

In July, Alphabet reported a 24% increase in revenue, reaching $119.8 billion. However, due to massive investments in AI infrastructure, the corporation recorded a negative free cash flow of $5.9 billion for the first time in its public history.

Previously, analysts from the Bank for International Settlements noted that the investment boom surrounding artificial intelligence, which supported the global economy in 2025, is itself becoming a source of macro-financial risks.

In a separate report, analysts warned about the financial risks of the AI boom. ForkLog also discussed how Bitcoin miners are shifting their infrastructure to support AI data centers.