This article first appeared on GuruFocus.
Meta Platforms (NASDAQ:META) is once again forcing investors to confront the central question behind the AI trade: how much spending is too much before Wall Street demands a clearer payoff. The company raised its full-year capital expenditure outlook to $125 billion to $145 billion, above analysts’ estimates and roughly 7.4% higher than its January forecast. Chief Financial Officer Susan Li said the increase reflects Meta’s conviction that its AI strategy is working, while also pointing to higher component pricing and additional data center costs. That matters because Mark Zuckerberg has already said Meta will spend hundreds of billions of dollars on AI infrastructure by the end of the decade, even before a memory chip shortage pushed prices higher. Since the beginning of the year, Meta has also announced billion-dollar deals with Nvidia (NASDAQ:NVDA), Advanced Micro Devices (NASDAQ:AMD), and Broadcom (NASDAQ:AVGO) for chips and other hardware, while building several massive data centers to support its AI push.
The results were not weak, but the spending narrative took control. Meta reported first-quarter sales of $56.3 billion, ahead of Wall Street’s $55.51 billion estimate, and guided current-quarter revenue to $58 billion to $61 billion, roughly in line with expectations. First-quarter net income reached $26.8 billion, including a one-time, non-cash income tax benefit of $8 billion tied to the implementation of the US tax policy signed into law in July. Analysts had estimated non-adjusted net income of $17.2 billion, without anticipating that benefit. Still, Zuckerberg’s confidence in heavier AI investment was not enough to calm investors. Shares fell as much as 7% in extended trading after he did not give detailed answers on how Meta plans to turn its AI spending into returns. The stock had closed at $669.12 in New York and was up 1.4% for the year.
The pressure point is that Meta’s AI strategy is getting larger while the path to monetization still looks unfinished. Bloomberg Intelligence analyst Mandeep Singh wrote that the higher spending increases the stakes for Meta because it is relying on its own AI system, which still trails frontier lab peers, and its standalone app has not generated the same engagement as other frontier labs. Meta’s daily active people across its platforms slipped to 3.56 billion in the first quarter, the first drop since the company began using that metric, with the company citing internet disruptions in Iran and Russia’s restrictions on WhatsApp access. Li said Meta would have seen positive user growth without those geopolitical issues. At the same time, the company is cutting costs, including roughly 8,000 jobs and leaving 6,000 open roles unfilled, with Evercore ISI estimating the May layoffs will save about $3 billion annually. Zuckerberg said Meta is already training more advanced models after debuting Muse Spark earlier in April, but also acknowledged he does not have a very precise plan for how each AI product will scale. For investors, that creates a sharp setup: Meta’s core business is still producing large revenue and profit, but AI infrastructure spending, user trends, and youth-related litigation could keep pressure on the stock until the company gives Wall Street a more convincing bridge from investment to returns.