That’s likely because the company raised its already lofty expense expectations for the year. Capital spending in 2026 will be at least $10 billion higher than expected and could exceed $145 billion, Meta said. CEO Mark Zuckerberg stressed his “confidence in this investment”, saying that much of this increase was due to “higher component costs, particularly memory pricing”.
The AI boom has led to an unprecedented data center buildout that has constrained global memory chip supply and driven up the prices of these valuable chips. The result has been a global memory crisis that has affected not only Meta and the rest of the AI industry, but also increased prices of consumer electronics like laptops and smartphones.
Meta’s $145 billion is a dramatic increase from the $72 billion capital spending recorded last year, and Zuckerberg is betting it all on the AI turnaround effort.
Meta has been left behind in the AI race as industry rivals like Google have moved ahead. About 10 months ago, Zuckerberg acknowledged the situation and announced a massive catch-up effort, under which he spent billions of dollars for research and development, and attracted talent from across the industry, including bringing in Scale AI founder Alexander Wang to lead the new Meta Superintelligence Labs AI division.
Many are understandably nervous about this commitment, given that the company’s latest big bet in emerging technology, the Metaverse, has flopped dramatically. In Wednesday’s earnings report, Meta said the Reality Labs division, which helped spearhead the Metaverse efforts, reported an operating loss of more than $4 billion, while sales generated only $402 million in cash. This has caused the division to lose $80 billion and more over the past six years.
But experts are somewhat optimistic about the AI bet because, earlier this month, the tech giant introduced the first fruits of that investment with AI model Muse Spark, a proprietary model that the company plans to open-source in the future. This is a step in the right direction, but META still needs to do more before it can confidently say that the catch-up effort is successful.
“This was the first release of Meta Superintelligence Labs, and it shows that our work is on the path to building a leading lab,” Zuckerberg assured investors in the company’s earnings call. “Now that we have a stronger model, we can develop even more new products.”
According to Zuckerberg, those new products will include two agents, one for personal use and the other for commercial use.
“We’re already testing an early version of Business AI and have seen weekly conversations increase 10x since the beginning of the year,” Zuckerberg said.
AI is clearly showing a way to benefit the meta internally. Meta CFO Susan Lee said more than half a billion users per week on Facebook and Instagram are now watching videos translated and dubbed by AI. The company is also incorporating new AI models into parts of its core business, such as ads and especially its recommendation system. The goal is to create AI hyper-personalized feeds for users.
“Because our recommendation systems are operating at such scale, we will phase in this new research and technology over time,” Zuckerberg said. “But it seems clear from the trends over the last few years that we are seeing increasing returns on the amount of money we can improve engagement for people and value for advertisers.”
AI is taking over internally in the meta as well. The company is laying off 10% of its workforce and reportedly offering voluntary buyouts to 7% of its US employees, which reportedly follows the AI-driven trend that has taken Silicon Valley by storm.
On the call, executives did not say whether the layoffs were related to the automation of jobs, but Lee said a “thin operating model” would “help offset the substantial investments we are making.”
<a href