Meta's next AI bet has one major catch for investors
Meta Platforms (META) has spent much of the artificial intelligence boom asking investors to trust the bill.
But now Wall Street might finally be getting a clearer answer on how Mark Zuckerberg plans to turn that spending into revenue.
Meta is establishing a cloud business to sell spare AI computing capacity, Reuters reported. The business might offer developers access to Meta's AI models or let clients purchase raw computing power, bringing Meta closer to the AI infrastructure market currently controlled by cloud giants and newer compute providers.
That's a big change for a corporation that still derives the bulk of its profit from digital ads.
Meta reported first-quarter revenue of $56.31 billion, $55.02 billion of which was from advertising. Its operating margin was 41%, a level few large technology companies can match.
So investors liked the cloud idea, but they may not be able to overlook the cost.
A cloud business might help Meta monetize its enormous AI and data-center buildout. But it also risks pulling the business into the lower-margin infrastructure market, where the economics are fundamentally different from Facebook and Instagram marketing.
Meta shares were recently trading at $582.90, giving the Facebook and Instagram parent a market capitalization of nearly $1.49 trillion.
Meta stock gets a new AI revenue story
The timing is important.
Meta has been spending big on AI infrastructure, processors and data centers, but investors want to know when that will translate into revenue.
The company stated capital expenditures, including principal payments on finance leases, were $19.84 billion for the first quarter. Meta also revised its 2026 capital expenditure outlook to between $125 billion and $145 billion, pointing to greater component prices and more data-center expenditures linked to future capacity.
Investors are more comfortable with that type of spending when there's a clear revenue stream tied to it.
A cloud business could provide one.
If Meta has more AI computing capacity than it needs for its models, ad tools, and consumer apps, selling that capacity to outside developers could make the buildout appear less like an unchecked cost and more like a platform business.
The idea also answers a broader strategic question for Meta.
Related: Meta just picked a fight with Amazon's cash cow
Most of the AI reward for Meta has so far been inside the advertising machine. AI helps enhance targeting, ad production and engagement across Facebook, Instagram and WhatsApp.
That's helpful, but it doesn't fully address the investor issue that Meta is spending tens of billions of dollars on infrastructure without building a new separate business.
Cloud computing could change that story.
Meta's cloud push could pressure margins
The catch is that cloud revenue is not ad revenue.
More Meta:
- Meta launches smart glasses cheaper than Ray-Ban
- Mark Zuckerberg admits mistakes in leaked memo after Meta layoffs
- Mark Zuckerberg and Meta face first tough test after layoffs
Meta's advertising business is unusually successful as the firm already owns the platforms, the audience and the auction system that sells ad space.
Cloud computing is another story because it takes massive infrastructure investments, enterprise clients, sales teams, service agreements, technical support, and ongoing investment in chips and data centers.
Alphabet (GOOGL) shows the contrast.
Google Services had $89.64 billion in sales and $40.59 billion of operating income in the first quarter. Google Cloud revenue was $20.03 billion, while operating income was $6.6 billion.
Key takeaways from Meta's cloud push
- Meta is reportedly building a cloud business to sell excess AI computing capacity.
- The move could help Meta monetize its heavy AI and data-center spending.
- Meta raised its 2026 capital-expenditure forecast to $125 billion to $145 billion.
- Advertising still accounted for nearly all of Meta's first-quarter revenue.
- Cloud computing could diversify revenue, but it may come with lower margins.
- Alphabet's results show cloud can be profitable, but the economics differ from ads.
- The investor question is whether Meta is selling spare capacity or entering a lower-margin infrastructure fight.
Google Cloud is a great business. It's a fast-growing business that is profitable presently.
But its profit profile is still distinct from the ad-heavy Google Services company. That's the problem Meta investors may have to start pricing in if cloud is to become a big part of the company's future.
Meta isn't concerned about whether cloud computing can make money.
The worry is that such revenue may compromise the margin profile that made Meta one of the most lucrative firms in tech.
Meta could put pressure on AI cloud stocks
Meta is not going to be the next Amazon Web Services overnight.
The more probable short-term course is more constricted: the sale of AI-specific processing capacity to developers and enterprises that need access to costly infrastructure.
That places the company in closer proximity to the world of AI-centric cloud vendors like CoreWeave (CRWV) and Nebius Group (NBIS) and not a full-service cloud behemoth.
The rumored Meta plan might put the company in competition with CoreWeave and Nebius, Reuters said.
That is why the report is significant outside Meta.
CoreWeave shares were last at $81.75, giving the business a market capitalization of around $43.1 billion. Nebius was currently trading at $215.62.
Meta has one advantage those companies do not.
Cloud does not have to be the whole story.
The same infrastructure may be leveraged by the corporation for its own AI models, ad products, and recommendation systems, as well as Meta AI, Instagram, Facebook and WhatsApp. Meta can offload any spare capacity. And if internal demand increases, Meta can consume more of it.
It's that flexibility that makes the strategy particularly compelling.
It also makes the margin question harder.
Meta's cloud push gives Wall Street what it wanted: a potential revenue stream directly tied to the company's AI spending.
But it also means investors have something fresh to worry about.
Meta's core ad business is asset-light relative to cloud infrastructure. Selling processing power would help justify the AI buildout, but it could also make Meta seem more like a capital-intensive infrastructure business on the fringes.
This is the true trade-off.
It could be a sensible approach to get more out of spending what it was already going to make if Meta can sell off idle AI capability without developing a large cloud operation.
If the company dives deeper into enterprise cloud, investors may have to accept a business with more revenue diversification but lower margins.
For now, Wall Street is a fan.
The next test is whether Meta can demonstrate that cloud computing is not simply a smart answer to AI spending concerns but a business that can increase revenue without eroding the profit profile that made the stock so attractive in the first place.
Related: Meta says it can read your thoughts without surgery
The Arena Media Brands, LLC THESTREET is a registered trademark of TheStreet, Inc.
This story was originally published July 4, 2026 at 2:47 PM.