Meta just secured Wall Street's approval for what might be the tech industry's most aggressive AI spending spree yet. The company's stock surged 10% in after-hours trading Wednesday after revealing plans to pour up to $135 billion into AI infrastructure in 2026—nearly double last year's capital expenditure. The market's enthusiastic response signals a pivotal shift: investors are willing to stomach massive AI investments as long as the core advertising business keeps printing money. With Q4 results showing 24% revenue growth, Mark Zuckerberg has earned himself an expensive blank check.
Meta just changed the rules on how much tech giants can spend chasing AI dominance. The company's Q4 earnings Wednesday didn't just beat expectations—they gave CEO Mark Zuckerberg permission to nearly double down on what's already the industry's most ambitious artificial intelligence buildout.
The numbers tell the story. Meta's revealing it will spend between $115 billion and $135 billion on AI-related capital expenditures this year, according to the earnings report. That's almost twice what the company invested in 2025, when it revamped its entire AI organization. Normally, that kind of spending announcement would send investors running. Instead, Meta shares popped as much as 10% in after-hours trading.
The reason? Meta's advertising engine is firing on all cylinders. Revenue grew 24% year-over-year, crushing analyst expectations and proving the company can fund its AI dreams without sacrificing profitability. It's the kind of performance that earns a CEO carte blanche to chase moonshots.
"As we plan for the future, we will continue to invest very significantly in infrastructure to train leading models and deliver personal super intelligence to billions of people and businesses around the world," Zuckerberg told analysts during the call. That "personal super intelligence" vision—his ambitious plan to build AI assistants that can understand and predict individual user needs—is now backed by Wall Street's full endorsement.
But Meta's facing a different problem than skeptical shareholders. The company is "capacity constrained," CFO Susan Li revealed on the earnings call. Translation: Meta can't build data centers and acquire computing power fast enough to meet internal demand. The company's AI team needs more resources to develop advanced models, while the advertising division requires additional compute to enhance targeting algorithms.
"Our teams have done a great job ramping up our infrastructure through the course of 2025, but demands for compute resources across the company have increased even faster than our supply," Li said. It's a high-class problem, but a real constraint on Meta's ability to execute its AI strategy at the pace Zuckerberg envisions.
The question investors should be asking: what exactly is Meta building with all this money? Zuckerberg's been notably vague about specific AI products that could generate meaningful new revenue streams. When pressed on the earnings call, he offered only generalities.
"I mean, we're going to roll out new products over the course of the year," Zuckerberg said. "I think the important thing is, we're not just launching one thing, and we're building a lot of things." That's not much detail for a $135 billion investment thesis.
The biggest bet so far is Meta's $14.3 billion acquisition of Scale AI last year, which brought founder Alexandr Wang and his team in-house. Wang now leads Meta's TBD AI unit, which has been testing a frontier model code-named Avocado designed to leapfrog the company's Llama family of models.
"I expect our first models will be good but, more importantly, will show the rapid trajectory that we're on," Zuckerberg said Wednesday. "And then I expect us to steadily push the frontier over the course of the year as we continue to release new models." Again, promising trajectory but light on specifics about commercial applications.
Zuckerberg defended the need for Meta to develop its own powerhouse foundation model by emphasizing the company's identity as a "deep technology company." Meta can't afford to be "constrained to what others in the ecosystem are building or allow us to build," he argued. Controlling your own model lets you "shape the future of these products" rather than depending on external providers like OpenAI or Google.
For now, online advertising still generates the overwhelming majority of Meta's revenue—and that's exactly why Wall Street is comfortable with the AI spending spree. As long as Instagram and Facebook keep dominating mobile advertising, beating quarterly expectations and throwing off billions in cash, Zuckerberg will have the runway to chase his AI vision.
The earnings beat essentially bought Meta's CEO another year of patience from investors. But at some point, that $135 billion has to produce AI products that do more than enhance ad targeting. The market's signaling it believes Zuckerberg can deliver. Now he has to prove it.
Meta's Q4 earnings just redefined how much tech companies can spend on AI without spooking Wall Street. The $135 billion capex plan is less a gamble and more a statement: if your core business is strong enough, investors will fund almost any AI vision. Zuckerberg's won his blank check, but the hard part starts now—turning massive infrastructure spending into products that justify the investment. With competitors like Google and OpenAI racing to build similar capabilities, Meta's capacity constraints might actually be the biggest risk. The company that moves fastest to turn compute power into breakthrough AI products will define the next era of tech. For now, Meta's advertising dominance has bought it time to find out if personal super intelligence is visionary or just expensive.
