Meta once defined scale by user count, but today ambition is measured in compute. Its third-quarter results showed a company racing to secure dominance in artificial intelligence as spending growth surpassed revenue gains. Years of cost control have shifted toward a capital surge typically seen in infrastructure buildouts.
Meta’s capital expenditures are projected to reach $70–$72 billion this year, with CFO Susan Li confirming a “notably larger” spend in 2026. AI now drives strategy at every level of Meta’s operations, shaping where capital flows and how teams are organized. CEO Mark Zuckerberg has made clear that the company will aggressively build data centers and compute to pursue superintelligence, a theoretical point where machines surpass human intelligence.
Funding scale
The company reported a 26 percent revenue rise in Q3 2025, but costs climbed 32 percent. Costs rose faster than revenue, making infrastructure the largest factor behind Meta’s expanding expenses. Meta’s lower-end capex forecast rose $4 billion from prior estimates. The company also recorded a $16 billion one-time charge tied to U.S. tax changes, pulling reported net income down to $2.71 billion. Without it, profit would have reached $18.64 billion.
Meta, Microsoft, Alphabet, and OpenAI are spending billions to build new data centers, driving record capital expenditures across the technology sector.
Talent density
Zuckerberg’s “Superintelligence Labs” unit, launched in June 2025, now concentrates the company’s most ambitious AI efforts. Meta is among the world’s largest buyers of Nvidia chips, and hiring remains intense. Employee compensation will rank as the second-largest cost contributor next year. CFO Li attributes the increase mainly to newly onboarded AI talent.
That talent concentration mirrors the capital intensity behind Meta’s expansion. The company’s approach combines compute acquisition with human capital scaling, aiming to secure both speed and expertise in the superintelligence race.
→ Explore more developments signaling industry disruption.
Investor pressure
Meta’s stock, up 28 percent this year, fell 8 percent after the earnings call. Investors are adjusting to a new reality where rising infrastructure costs may take years to pay off. Analysts point to a growing tension between near-term returns and the scale required for AI dominance.
The $70 billion threshold aligns Meta’s capital plan with the scale of national energy projects and large infrastructure initiatives. It represents one of the largest corporate investments in compute capacity ever recorded. That level of ambition increases pressure to demonstrate efficiency and sustained performance.
Market positioning
Despite the expense, Meta’s advertising engine remains powerful. The company’s AI-optimized ad tools are sharpening campaign automation, improving video quality, and expanding reach across Instagram Reels, WhatsApp, and Threads. More than 3.5 billion people use at least one Meta app daily, reinforcing the platform’s dominance.
For now, AI directly supports the company’s core business. Even if superintelligence remains distant, the compute built for it enhances ad performance, audience targeting, and product development.
Near-term outlook
Meta’s aggressive capital allocation hints at a longer horizon for profitability, but also at deeper integration of AI across its stack. The company’s forecasted Q4 revenue range of $56–$59 billion aligns with steady ad growth but implies limited short-term margin relief. The question now is whether Meta can sustain investor confidence through the superintelligence buildout cycle.
If this trajectory continues, infrastructure leadership will become the new benchmark for competitive advantage. Meta’s challenge is turning that capacity into measurable returns before market patience runs out.
Reference
Singh, J., & Wang, E. (2025, October 29). Meta forecasts bigger capital costs next year as Zuckerberg lays out aggressive AI buildout. Reuters. https://www.reuters.com/business/metas-profit-hit-by-16-billion-one-time-tax-charge-2025-10-29/
