Meta Pivots to Cloud Infrastructure to Monetize Massive AI Spending

The company’s plan, internally referred to as Meta Compute, places it in direct competition with established cloud giants like Amazon Web Services, Google Cloud, and Microsoft Azure. By offering external access to its hardware and proprietary models—including the recently launched Muse Spark—Meta hopes to transform its massive internal projects in Ohio and Louisiana into a revenue-generating business line. This initiative is spearheaded by infrastructure head Santosh Janardhan, Daniel Gross of Meta Superintelligence Labs, and president Dina Powell McCormick.

While Meta has invested heavily in AI, its own models and services have yet to prove their value as standalone revenue drivers. The move reflects a broader trend in the tech industry where firms are increasingly looking to lease out excess capacity to offset the immense costs of building out AI-ready facilities. However, the strategy arrives amid growing skepticism regarding the long-term sustainability of the AI infrastructure boom, with some market observers questioning whether end-user demand will eventually justify these trillion-dollar expenditures on rapidly depreciating hardware.