Meta plans cloud business to sell AI compute and models

Meta is developing plans for a cloud infrastructure business that would sell access to AI computing power and models, a move that would push the company into the same market as Amazon Web Services, Google Cloud and Microsoft Azure. The effort sits inside an internal initiative called Meta Compute and is being led by Meta infrastructure chief Santosh Janardhan, Daniel Gross of Meta Superintelligence Labs and Meta president Dina Powell McCormick. A Meta spokesperson declined to comment.

The timing lands as Big Tech races to turn expensive AI buildouts into revenue. Meta shares rose more than 6% in premarket trading after the plan became public, a sign that investors see a chance for the company to monetize the data centers, chips and networking it has been pouring into AI. The push also reflects a broader scramble across the industry to find returns on infrastructure spending that has become massive, and in some cases difficult to justify as a purely internal cost.

The shortage of compute is already shaping business decisions. Reuters reported on June 28 that Google had put limits on Meta’s use of Gemini AI models after Meta sought more computing capacity than Google could provide. That constraint underscored how tightly held advanced compute has become, even among the largest tech firms, and why Meta may want to create a market for any excess capacity it controls rather than rely on outside providers.

The opportunity is not hypothetical. On May 6, Anthropic said it had reached an agreement with SpaceX to use all of the compute capacity at the Colossus 1 data center in Memphis, Tennessee. Anthropic said the deal gave it access to more than 300 megawatts of capacity and over 220,000 NVIDIA GPUs, and that the added SpaceX resources would directly improve access for Claude Pro and Claude Max subscribers. That arrangement shows how AI infrastructure is starting to look like a utility business, with compute sold in bulk to firms that need more power than they can build quickly on their own.

For Meta, a cloud business would not just add another line of revenue. It would turn the company’s own infrastructure into a product and deepen the competitive pressure on the cloud giants that already dominate enterprise computing, while also concentrating even more control over AI capacity in a handful of powerful firms.