Meta Platforms is reportedly planning to launch a cloud infrastructure business aimed at selling access to AI computing power and models. This strategic move would place Meta in direct competition with established industry leaders such as Amazon Web Services, Microsoft Azure, and Google Cloud. The initiative is intended to generate revenue from the company's significant investments in AI infrastructure, including expensive data centers.
According to sources familiar with the matter, Meta's new business line, potentially named Meta Compute, will be led by Santosh Janardhan (Head of Infrastructure), Daniel Gross (Meta Superintelligence Labs leader), and Dina Powell McCormick (Meta President). The company is considering multiple approaches, including selling access to various AI models hosted on its infrastructure, such as Muse Spark, similar to AWS's Bedrock offering. Additionally, Meta might sell raw computing capacity, mirroring the business model of "neocloud" providers like CoreWeave.
This development comes as Meta is projected to spend up to $145 billion on AI infrastructure this year, part of a larger trend of significant investment by tech giants into AI. CEO Mark Zuckerberg had previously indicated that entering cloud computing was "definitely on the table," noting that companies frequently approached Meta to purchase access to its AI models or spare computing power. While Meta has not officially confirmed these plans, the news led to a rise in Meta shares and a decline in shares of neocloud companies like CoreWeave and Nebius, due to concerns about increased competition and reduced reliance on their services by Meta.
Analysts suggest this move could help Meta tap into growing demand for AI services and reduce its dependence on advertising revenue. However, it also raises questions about Meta's focus on catching up with leading AI labs. The company’s plans are still in development, and the strategy could change.
