Meta raises 2026 capex to $145B as AI spending doubles year-over-year

Meta Platforms just told the market it plans to spend more money on AI than most countries collect in annual tax revenue. The company raised its full-year 2026 capital expenditure guidance to a range of $125B to $145B, up from a prior forecast of $115B to $135B. Investors were not thrilled.

Shares dropped roughly 10% following the April 29-30 earnings release, erasing approximately $150B in market capitalization.

What actually changed

The headline number is striking on its own, but the year-over-year math is what puts it in perspective. Meta’s actual 2025 capital expenditures came in at $72.2B. The midpoint of the new 2026 guidance sits at roughly $135B, which works out to an 87% increase year-over-year. In English: Meta is planning to spend nearly twice as much building AI infrastructure this year as it did last year.

The company cited two main drivers. First, the cost of AI compute components has risen faster than originally modeled. Second, internal demand for those components has grown beyond earlier projections.

Meta also announced plans to cut roughly 8,000 employees, representing approximately 10% of its workforce. The layoffs are designed to keep overall operating expense forecasts intact even as capital spending surges.

Where this fits in the broader hyperscaler arms race

Meta is not alone in this territory, but it is moving with unusual aggression. Combined 2026 capital expenditure estimates across major hyperscalers have crossed $700B, with Amazon, Google, and Microsoft all running their own AI infrastructure buildouts. Amazon, notably, did not raise its guidance on its own Q1 print, which makes Meta’s upward revision stand out even more in comparison.

For Meta specifically, the AI infrastructure is not just about competing with OpenAI or Anthropic in the model race. The more immediate goal is improving the AI systems embedded in Facebook, Instagram, WhatsApp, and the ad products that sit behind all of them. Meta has roughly three billion people using its apps every day.

What this means for investors watching the sector

The market’s reaction to Meta’s earnings is a useful data point for anyone trying to read investor sentiment on AI spending broadly. A 10% single-day drop at a company of Meta’s size reflects genuine uncertainty about the timeline for returns on capital being deployed at a pace that has no real precedent in corporate history.

Meta’s stock had been performing well heading into the print, which means the guidance raise, paradoxically, became a negative catalyst. Investor apprehension centered specifically on concerns regarding expected return on investment from the escalated AI spending.

Longer-term investors will be watching whether Meta can convert this infrastructure into measurable improvements in ad revenue efficiency and user engagement metrics over the next several quarters.

Disclosure: This article was edited by Editorial Team. For more information on how we create and review content, see our Editorial Policy.