Meta jumps after report says it is weighing a plan to sell AI compute

Meta Platforms shares surged in morning trading after a report circulated about potential plans to sell AI compute to outside customers, a move that would put the company closer to the core infrastructure business dominated by cloud and data-center providers.

The steep gain, described as about 9% to roughly $613, reflects how quickly investors can react to any prospect of new revenue streams tied to artificial intelligence infrastructure. The catalyst for the move was a Bloomberg report that Meta is “building a…” related initiative, according to the account driving the market chatter.

If Meta’s intent is to offer AI compute capacity to third parties, it would represent a shift from Meta’s mostly internal use of servers and GPUs to support its own models and ad targeting. In general terms, AI compute can be thought of as the processing power needed to train and run machine-learning systems, and it is typically delivered through managed cloud services.

The same report framing, as repeated by the market coverage, suggests Meta’s potential positioning would challenge large incumbents that already supply AI and cloud capacity: Amazon’s AWS, Microsoft’s Azure and Google’s cloud stack. In the broader market, these players sell compute through subscriptions and usage-based pricing, and they compete on performance, availability, and integration with AI tooling.

The reaction also underscores how the AI buildout has become a business strategy, not just a technology race. Companies that can convert data-center scale into sellable services can diversify beyond advertising, which is often the primary cash engine for social and internet platforms.

For Meta, the practical question investors will likely have is what exactly would be sold, to whom, and under what commercial terms. In the coverage that drove the move, key specifics about product scope and customer targeting were not laid out, and the reporting cited in the market item did not provide figures on expected capacity, pricing, or timeline.

Alphabet and its peers matter in this story because they are already competing for enterprise AI workloads, whether through cloud platforms or AI services layered on top of underlying infrastructure. If Meta were to move from internal infrastructure to external offerings, it would widen the set of vendors enterprises could consider when procuring AI compute.

Still, it is not clear from the available market coverage what Meta has formally decided, what approvals or partnerships would be required, or how quickly any sellable offering could scale. Until the company provides additional detail in a filing, earnings communication, or direct announcement, the market impact may be driven more by possibility than by a defined, near-term business plan.