Meta makes cloud push to sell excess AI compute, following SpaceX’s playbook as Big Tech seeks returns on AI spending

For the past two years, Big Tech has poured hundreds of billions of dollars into AI chips, data centers, and computing infrastructure with one goal in mind: to build enough capacity to stay ahead in the AI race. Now, a new phase is emerging. Instead of keeping all that infrastructure behind closed doors, companies are looking for ways to turn spare computing power into a business.

Meta is the latest to make that move.

The parent company of Facebook and Instagram is developing a cloud business that would let developers rent access to its AI infrastructure and models, according to a Bloomberg News report published Wednesday. The effort remains under development, and the strategy could still change, according to people familiar with the plans.

“Meta Platforms Inc. is developing plans for a cloud infrastructure business that will sell access to AI computing power and models, setting up a new vector of competition with industry leaders like Amazon Web Services, Microsoft Azure and Google Cloud,” Bloomberg reported.

The move mirrors a shift already underway across the AI industry. Earlier this year, SpaceX began renting access to computing capacity at its Memphis AI data center through agreements with Anthropic and Google. Rather than letting expensive infrastructure sit idle, companies are discovering they can generate new revenue by selling unused AI compute.

Investors welcomed the report. Meta shares climbed more than 10% on Wednesday, easing pressure on a stock that had fallen nearly 15% this year through Tuesday and trailed the broader S&P 500, according to CNBC.

The news weighed on AI infrastructure providers that have benefited from Meta’s spending spree. Shares of CoreWeave dropped 10.8%, and Nebius fell 12.4%, reflecting concerns that Meta could rely less on outside cloud providers if it begins selling computing capacity from its own infrastructure.

Why Meta wants to sell AI compute instead of keeping it idle

The strategy could place Meta in more direct competition with cloud giants Amazon, Microsoft, and Google, all of which already rent AI infrastructure and model access to businesses. Bloomberg reported that Meta’s planned service would allow developers to run AI models hosted on Meta’s infrastructure, including its Muse Spark model, paying for the computing resources required to use them. The company is reportedly evaluating a second offering to sell raw AI computing capacity, similar to services offered by AI-focused cloud companies such as CoreWeave.

The approach closely resembles Amazon Web Services’ Bedrock platform, which provides developers with access to AI models from multiple providers via AWS infrastructure. Meta has yet to announce when such a service could launch.

The cloud push arrives as questions continue to surround Meta’s AI strategy. Chief Executive Mark Zuckerberg has spent billions rebuilding the company’s AI organization, recruiting researchers from competitors and investing heavily in infrastructure. In April, Meta introduced Muse Spark, the first model developed by its new AI team, though it has yet to release the model publicly. The Wall Street Journal reported last month that no launch date has been scheduled.

Analysts say the biggest impact may fall on AI cloud providers rather than traditional hyperscale cloud companies.

“The impact of adding Meta’s capacity to the market is more likely to be on neoclouds than the big hyperscalers. Those companies like CoreWeave and Nebius rely on Meta for their growth, and Meta may not need them anymore,” said Gil Luria, managing director at D.A. Davidson.

Billions spent on AI infrastructure are creating a new cloud business opportunity

The economics behind the decision are becoming harder to ignore. Meta is expected to spend up to $145 billion on AI infrastructure this year, part of a broader wave of more than $700 billion in AI spending across the largest technology companies. That level of investment has increased pressure on executives to find new revenue streams beyond advertising or consumer subscriptions.

Zuckerberg has already hinted that cloud computing could become part of Meta’s business. Speaking during the company’s shareholder meeting in May, he said entering the cloud market was “definitely on the table,” adding that companies were approaching Meta “almost every week” seeking access to its AI models and spare computing capacity.

That demand is starting to reshape how the industry’s biggest infrastructure owners think about AI. The race is no longer limited to building larger GPU clusters. It is increasingly about making those multibillion-dollar investments pay for themselves.