Meta has invested billions into artificial intelligence and the data centers needed to power it. Now, the company appears ready to put those facilities to work generating more immediate revenue. According to a Wednesday report from Bloomberg, Meta is working on plans to launch a cloud infrastructure business, offering access to both AI computing power and AI models. This would put the company in direct competition with major cloud providers such as Amazon Web Services, Google Cloud, and Microsoft Azure.

Meta’s push to sell its surplus computing capacity comes just weeks after SpaceX, through its xAI division, announced a similar initiative. In early May, SpaceX struck a deal with Anthropic to purchase all the compute capacity at SpaceX’s Colossus 1 data center. SpaceX has also signed additional leases with Google and Reflection AI. The fact that Meta is following the same path suggests the real winners in the AI race may not be those with the best models or services, but rather those who control the data centers themselves—assuming demand for computing power stays strong and those data centers hold their value.

Some skeptics have warned that the rush to build AI infrastructure is creating a bubble, heavily dependent on chips that depreciate rapidly. Others question whether AI companies can generate enough end-user revenue to justify trillion-dollar investments. These concerns have not slowed Meta’s commitment to AI infrastructure. As of the end of the first quarter, Meta had pledged to spend $182.9 billion on AI infrastructure over the coming years, including massive ongoing projects in Louisiana and Ohio. The Ohio facility, which Mark Zuckerberg described as being the size of Manhattan, is expected to become operational this year.

Unlike Google and OpenAI, Meta has not seen significant demand for its own AI models and services. The company does not break out revenue from Meta AI or Llama, its family of open-weight AI models, in its earnings reports. Executives have largely emphasized the internal corporate uses of AI in public statements, suggesting that Meta’s AI efforts may not yet represent a material standalone revenue stream.

To generate a return on its enormous spending, Meta may adopt CoreWeave’s business model and sell access to "raw" computing capacity, according to Bloomberg. The outlet also reported that Meta is considering following AWS’s lead by offering access to various AI models—including its recently launched closed-weight model, Muse Spark—hosted on its AI infrastructure. This new business line will reportedly fall under an initiative called Meta Compute, led by head of infrastructure Santosh Janardhan, Meta Superintelligence Labs leader Daniel Gross, and president Dina Powell McCormick. The report confirms statements Zuckerberg made in May, when he said a Meta cloud computing business is "definitely on the table" as a way to recoup some of the massive investment tied to his strategy of developing AI "superintelligence."