Google assumes $44 billion in data center lease risks

What happened?

Google has disclosed in its latest financial reports that it has assumed lease guarantee commitments worth up to $44 billion for data centers it does not own. The figure, reported by The Next Web and confirmed by The Information, marks a spectacular jump from $6.5 billion recorded at the end of September 2024. This nearly sevenfold increase in less than a year reflects Google's urgency to secure computing capacity for its AI models, such as Gemini, and for its Google Cloud customers, including startup Anthropic, in which it has invested over $2 billion.

These guarantees function as insurance: if the primary tenant (usually an investment fund or data center operator) cannot pay the rent, Google commits to covering it. In return, Google gains priority access to computing capacity without having to build or directly finance the facilities. This model is similar to that used by other tech giants like Microsoft and Amazon, but Google's scale is particularly striking.

Why is this important?

This move reflects several key trends:

• Acceleration of AI investment: Google needs more computing capacity to train and run models like Gemini, as well as to serve Google Cloud customers and partners like Anthropic. According to company data, capital expenditure on technology infrastructure reached $32 billion in 2024, and is expected to exceed $50 billion in 2025.
• Outsourcing real estate risk: By not owning the facilities, Google avoids the massive capital expenditure of building data centers but assumes significant contingent risk. This practice has become common among hyperscalers: Microsoft, for example, reported lease commitments of over $30 billion in its latest financial report.
• Competitive pressure: Microsoft and Amazon are also investing aggressively in cloud infrastructure. Microsoft has announced plans to spend $80 billion on data centers in 2025, while Amazon Web Services (AWS) has committed $150 billion over the next 15 years. Google's figure, though smaller, is comparable in terms of relative growth.

The magnitude of the commitment ($44 billion) is comparable to the GDP of countries like Slovakia or Luxembourg, and exceeds the market value of many tech companies. For perspective, Uber's market capitalization is around $80 billion, Airbnb's about $70 billion. This level of financial leverage is unusual even for a company the size of Alphabet.

Consequences and risks

If demand for cloud services slows or if primary tenants default, Google could face massive payments. However, the company is confident that AI growth will keep demand high. Analysts note that this strategy is similar to what Google used in the past with its office lease agreements, but on a much larger scale. During the dot-com bubble, many tech companies took on rental commitments they couldn't sustain, leading to bankruptcies. But Google has a solid balance sheet: $110 billion in cash and equivalents, allowing it to absorb potential losses.

The risk is not just financial. Massive data center construction has energy and environmental implications. According to the International Energy Agency, data centers consume about 1% of global electricity, and this figure is expected to double by 2026 due to AI. Google has committed to operating on carbon-free energy by 2030, but its lease guarantees do not include explicit sustainability clauses, which could draw criticism.

Additionally, these agreements could distort the data center real estate market. By acting as a guarantor, Google drives up rental prices and makes it harder for smaller companies to compete. The Information reports that some data center operators have seen their leasing costs surge 30% in the past year, partly due to hyperscaler demand.

What should readers know?

For investors, this data signals that Google is betting big on AI, but also adds financial risk worth monitoring. Credit rating agencies like Moody's and S&P have already noted that rising contingent commitments could affect Alphabet's rating if the trend continues. For now, they maintain a stable outlook but warn that further increases could lead to a review.

For businesses using Google Cloud, the news is positive: it indicates Google is expanding capacity to offer AI and cloud computing services, translating to lower latency, higher availability, and potentially more competitive pricing. However, it could also mean Google will pass some costs to customers in the medium term.

For the tech sector overall, it confirms that the AI infrastructure race is intensifying, with implications for energy consumption, chip supply chains, and data center real estate. Companies like NVIDIA, which manufactures the GPUs needed for AI, directly benefit from this demand. In fact, NVIDIA reported record revenue of $60 billion in 2024, largely driven by hyperscalers.

In summary, Google's $44 billion in lease guarantees is not just a striking number: it represents a strategic bet that redefines the balance between risk and opportunity in the AI era. The coming years will show whether this outsourcing of real estate risk is a masterstroke or a financial burden.