When Meta reportedly turned to JPMorgan Chase and Morgan Stanley to help raise roughly $13 billion for a new Texas data centre project, the move signalled towards how the economics of the artificial intelligence race is evolving. 

For years, Big Tech companies funded data centre growth largely through operating cash flows generated by cloud computing and advertising. But the rise in spending on AI—which includes infrastructure such as massive number of GPUs, custom chips, networking systems and hyperscale data centres, are leading to some of Silicon Valley’s richest companies increasingly leaning on debt markets and external financing.

What's interesting to note is the timing of Meta’s financing discussions. Just days earlier, the company raised its 2026 capital expenditure guidance to between $125 billion and $145 billion, up from its prior forecast of $115 billion to $135 billion, citing it as AI infrastructure investments.  

But unlike rivals such as Microsoft, Amazon and Alphabet, Meta lacks two important buffers supporting the current AI boom—a large cloud computing business monetising AI demand directly, and lucrative equity stakes in external AI leaders like Anthropic or OpenAI.

That distinction is beginning to define how investors are interpreting Meta’s AI strategy.

How rivals are monetising the AI boom

The latest earnings calls from the Big Tech companies showed that AI demand is already boosting cloud businesses.

Alphabet reported that Google Cloud revenue surged 63% year-on-year to $20 billion in the first quarter, with CEO Sundar Pichai stating that the company’s “Enterprise AI solutions have become our primary growth driver for Cloud for the first time.”  

The company also raised its 2026 capital expenditure guidance to $180 billion–$190 billion, while CFO Anat Ashkenazi said Alphabet was seeing “unprecedented internal and external demand for AI compute resources.”  

Meanwhile, Amazon and Microsoft continue to benefit from AI-driven demand flowing through Amazon Web Services and Azure, with both companies positioning cloud infrastructure as the foundation of the generative AI boom.

But cloud revenue is only part of the story. Another, less discussed contributor to recent profits has been the sharp rise in valuations of external AI startups. Amazon disclosed a $16.8 billion pre-tax gain linked to its investment in Anthropic as the startup’s valuation surged. Alphabet, which holds around 15% stake in Anthropic noted a $36.9 billion equity gain driven by the surging value of its stake. In Microsoft's case, it's adjusted earnings exclude a $14 million decrease in net income from OpenAI investments.

These gains matter because they create a second layer of AI monetisation beyond infrastructure revenue. 

In short, these companies are making money from cloud revenue from AI demand, investment upside from AI startups, and their own internal AI products and models. This diversification helps cushion the AI spending. 

Meta’s AI strategy stands alone

Here's where Meta’s position is different. The company has no large cloud division selling AI infrastructure to enterprise customers. It also lacks major equity stakes in frontier AI startups that could generate investment gains as the market expands.

Yet the company is spending at a scale comparable to hyperscalers. In the first quarter alone, the company reported capital expenditure of $19.8 billion, driven by investments in “servers, data centres, and network infrastructure.” It also disclosed that multi-year cloud deals and infrastructure purchase agreements have led to a $107 billion increase in contractual commitments during the quarter.  

Unlike rivals, however, Meta’s infrastructure spending is largely designed to support its own ecosystem of recommendations, advertising products, AI agents, and internal model development.

That distinction emerged clearly during the company’s earnings call, where analysts repeatedly pressed executives on return on investment.

Analysts asked CEO Mark Zuckerberg what were the key factors that Meta was watching to ensure it could “generate healthy ROIC on all this CapEx and infrastructure spend.”  

Zuckerberg’s response reflected Meta’s long-standing product philosophy. “The formula for our company has always been building experiences that can get to billions of people and focus on monetizing them once you get to scale,” he said.  

That strategy worked successfully for Facebook, Instagram, and WhatsApp. But applying the same “scale first, monetise later” playbook to AI infrastructure spending which involves billions creates a very different level of financial risk.

The uncertainty surrounding AI demand also surfaced during the call. Asked about future spending, Meta CFO Susan Li acknowledged that the company had “continued to underestimate our compute needs even as we have been ramping capacity significantly.” This suggests that even Meta itself does not know how much infrastructure its AI ambitions will ultimately require.

Meta is also trying to reduce long-term dependence on Nvidia by developing custom AI chips with Broadcom. But unlike Amazon’s Trainium chips or Google’s TPUs, which are tied to cloud services sold externally, Meta’s chips will be used to increase efficiency for internal workloads.

Simply put, it means that Meta’s infrastructure investments are not directly generating revenue in the same way as cloud competitors.

Is AI becoming a financing story

Meta's discussions with JPMorgan for financing and frontier AI startups partnering with PE firms seem to point towards the fact that while there's neck-to-neck competition in the tech race, there's also a race for financing.

Data centres are becoming utility-scale assets requiring long-term financing structures more commonly associated with infrastructure projects than software companies.

Companies such as Amazon, Microsoft and Alphabet can spread AI spending across cloud customers, enterprise contracts and broader AI ecosystems. Meta, by contrast, is largely relying on improvements to advertising, engagement and future AI products to justify its infrastructure buildout.

For now, Meta remains to be making an AI bet at hyperscale without the safety net of a cloud business or large investment windfalls from external AI startups.