Meta's massive artificial intelligence infrastructure spending is no longer a guaranteed Wall Street win. The company reported strong earnings in Q1 2026, with revenue jumping 33 percent to $56.31 billion and beating analyst expectations on both top and bottom lines. But when Chief Financial Officer Susan Li announced that 2026 capital expenditures would climb to between $125 billion and $145 billion, up from the prior range of $115 to $135 billion, investors reacted with alarm. The stock fell 7 percent in after-hours trading and dropped another 3 percent the following day, erasing roughly $175 billion in market capitalization across two trading sessions.

Why Did Wall Street Turn on Meta's AI Strategy?

The downgrade from JPMorgan analyst Doug Anmuth captured the shift in investor sentiment. Anmuth cut his rating from Overweight to Neutral and slashed his price target from $825 to $725, tying the move directly to Meta's escalating capital intensity. His concern was not about the spending itself, but about whether Meta could actually generate returns from it. Unlike Alphabet, which raised its 2026 capex guidance to between $180 billion and $190 billion on the same day and saw its stock rise 7 percent, Meta has no direct revenue line item tied to AI infrastructure investments.

The difference between the two companies illustrates a critical distinction Wall Street is now making. Alphabet's Google Cloud division grew 63 percent year-over-year to $20 billion in revenue, with a backlog that roughly doubled to $460 billion. Investors can see a direct path from infrastructure spending to revenue growth. Meta's $125 to $145 billion capex commitment, by contrast, funds advertising algorithms, Reels recommendation models, and pre-training runs for future models developed by Meta Superintelligence Labs. The returns are supposed to materialize through better ad targeting and engagement on Facebook, Instagram, and WhatsApp, but that connection is less tangible to investors.

"While we're encouraged by META's 33 percent year-over-year revenue growth which has been supported by AI-driven ad stack optimizations, accelerating impression growth, and engagement gains, we believe full-stack AI competition is intensifying and Meta has a more challenging path to returns on heavy AI capex beyond advertising," stated Doug Anmuth, analyst at JPMorgan.

Doug Anmuth, Analyst at JPMorgan

What Do Meta's Own Numbers Reveal About the Spending Trajectory?

The scale of Meta's infrastructure buildout becomes clearer when examining the company's actual quarterly spending. In Q1 2026 alone, Meta deployed $19.8 billion in capital expenditures, primarily across servers, data centers, and network infrastructure. Annualized, that quarterly pace would already push the company past the low end of its full-year guidance. Mark Zuckerberg told analysts that Meta is now rolling out more than one gigawatt of its own custom silicon designed with Broadcom, alongside a significant amount of AMD chips, in an effort to reduce dependence on Nvidia GPUs and bring per-token training costs down.

The most troubling signal came when Susan Li declined to provide any 2027 capex outlook at all. She told analysts the company is "frankly undergoing a very dynamic planning process" and acknowledged that Meta has "continued to underestimate" its compute needs even as it has ramped capacity. That admission, that the spending bill keeps growing faster than Meta's own forecasts, is what triggered Anmuth's downgrade. His model now projects Meta capex jumping another 42 percent to $202 billion by 2027, which would result in negative free cash flow of roughly $4 billion in 2026 and $24 billion in 2027, the first such gap in over a decade.

She

How Has Meta's AI Spending Escalated Over Time?

• FY 2024: Meta deployed $39.2 billion in capital expenditures, representing the last "normal" capex year before the generative AI buildout took over the budget.
• FY 2025: Meta's capex roughly doubled to $72.2 billion, reported in the January 2026 Q4 earnings release, signaling the beginning of the infrastructure acceleration.
• January 2026 Guidance: Meta initially guided for 2026 capex between $115 billion and $135 billion, framing the year as a deliberate front-loading of compute capacity.
• April 2026 Revision: Susan Li raised the 2026 range by $10 billion on each end to $125 billion and $145 billion, with Q1 capex alone reaching $19.8 billion.
• Analyst Projections: JPMorgan now projects 2027 capex could reach $202 billion, representing a 42 percent year-over-year increase.

This escalation pattern reveals a fundamental challenge facing Meta and other hyperscalers. Each time the company revises its capex guidance upward, it signals that prior estimates were too conservative. The market interprets these revisions as evidence that management cannot accurately forecast compute requirements, raising questions about whether the company truly understands the return on investment for its infrastructure spending.

What Are the Broader Implications for Meta and Big Tech?

The JPMorgan downgrade carries specific signals for different constituencies within and around Meta. For machine learning engineers and infrastructure teams inside the company, the downgrade reinforces pressure on operating expenses. Meta's planned May 20 layoffs of roughly 8,000 workers were already structured as headcount reduction in service of AI infrastructure, but with a $145 billion capex ceiling now formal and a $202 billion 2027 trajectory becoming the new analyst baseline, the financial math will keep pressing on operational spending. AI Foundations, the Reality Labs cost center, and Llama-era applied research teams are the most exposed to potential cuts.

For practitioners at Anthropic, Google, OpenAI, and other companies that Meta buys compute and licenses from, the takeaway is more ambiguous. The capex still gets spent in 2026, so Meta is not retreating from its current budget. But the next leg of the buildout, the leg that funds 2027 and 2028 compute, is now being underwritten under explicit Wall Street skepticism for the first time. Google's $40 billion deal with Anthropic, announced the week before Meta's earnings, was structured precisely to shift some of that risk onto a model partner.

Not every Wall Street desk turned bearish on Meta. Several analysts left their Buy ratings unchanged and pointed to the same Q1 numbers JPMorgan dismissed. Daily active users on Meta's family of apps grew, Reels monetization continued accelerating, and Susan Li reiterated that Meta's 2026 expense growth would land roughly in line with prior expectations. The divergence in analyst opinion reflects genuine uncertainty about whether Meta's AI infrastructure spending will eventually pay off through improved advertising performance or whether the company is overcommitting capital to a strategy with uncertain returns.