Meta Challenges Amazon and Google with Ambitious AI Cloud

Meta Challenges Amazon and Google with Ambitious AI Cloud

Meta has invested billions of dollars in the development of AI and the expansion of its data centers to support it. However, the company now seems ready to utilize these data centers for more immediately profitable purposes.

On Wednesday, Bloomberg reported that Meta is crafting plans for a cloud infrastructure service, selling access to both AI computing power and models. This initiative would put it in competition with major cloud providers such as Amazon Web Services, Google Cloud, and Microsoft Azure.

Meta's decision to sell its excess computing capacity comes just weeks after SpaceX, through xAI, announced similar plans. In early May, SpaceX signed an agreement with Anthropic to acquire all the computing capacity of its Colossus 1 data center. SpaceX has since signed similar leases with Google and Reflection AI. The fact that Meta is following suit indicates that the winners in the AI race may not be those providing the best models and services, but rather those who own the data centers.

That said, it will depend on the demand for computing and the value of the data centers. Some skeptics have warned that the race to build AI infrastructure is creating a bubble heavily reliant on chips that depreciate rapidly. Others have questioned whether AI companies can generate enough revenue from end users to justify trillion-dollar bets.

These concerns have not deterred Meta from massively investing in infrastructure for AI computing. By the end of the first quarter, Meta had committed $182.9 billion for AI infrastructure in the coming years, including significant projects in Louisiana and Ohio. The project in Ohio, which Zuckerberg described as being the size of Manhattan, is expected to be operational this year.

Unlike Google and OpenAI, Meta has not seen significant demand for its own AI models and services. Meta does not detail its revenues from Meta AI or Llama, its family of open-weight AI models, in its financial results, and executives have primarily highlighted internal uses of AI in their public statements. This could mean that Meta's AI efforts do not yet represent a significant standalone revenue line.

To recoup some of its colossal investments, Meta might copy the business model of CoreWeave and sell access to "raw" computing capacity, according to Bloomberg. The agency also reported that Meta is considering following AWS's example and selling access to various AI models—including its recently launched closed-weight model, Muse Spark—hosted on its AI infrastructure.

This new line of business will be part of a new initiative seemingly dubbed Meta Compute, led by infrastructure head Santosh Janardhan, Meta Superintelligence Labs leader Daniel Gross, and president Dina Powell McCormick.

The report confirms Zuckerberg's statements in May that a Meta cloud computing venture is "definitely on the table" as a means to recoup some of the massive investments in its "superintelligent" AI development strategy.