Meta capex inflation is accelerating — and the numbers are staggering. The company formerly known as Facebook is now projecting up to $72 billion in capital expenditures, fueled by Mark Zuckerberg’s aggressive bet on AI infrastructure and lingering cost pressure from Trump-era tariffs.
The revised spending guidance, revealed in Meta’s latest earnings update, reflects both ambition and necessity. Zuckerberg has made AI the company’s top investment priority, committing to building world-class compute capacity — including data centers, energy contracts, and proprietary chips.
But inflationary pressures are multiplying. As seen in Millionaire MNL, infrastructure projects are being hit by higher material prices, skilled labor shortages, and tariffs that increase the cost of importing critical components like servers and networking gear.
AI growth isn’t cheap
Zuckerberg has called AI “the most important technology of our time” and has vowed to embed it deeply across Meta’s products — from Instagram Reels recommendation engines to WhatsApp chat assistants and business tools.
To support that vision, Meta is building new data centers at a rapid pace, upgrading compute clusters to run large language models (LLMs), and reportedly developing its own custom AI chipsets.
All of that comes at a steep cost. According to Meta’s CFO, infrastructure spending alone could exceed $35 billion this year, with the rest allocated to long-term hardware, research, and energy contracts.
The Meta capex inflation spike highlights the broader economic reality: AI scale-up requires real-world assets, and building them in a high-cost environment is straining even the biggest balance sheets.
Trump tariffs add a new layer of friction
While much of the AI discussion has focused on software breakthroughs, the physical layer of AI — servers, power supplies, GPUs — is vulnerable to trade policy shifts.
Trump-era tariffs on Chinese-made tech components remain in effect, and while the Biden administration has explored adjustments, many duties persist. Meta sources a large portion of its networking and server hardware from Asian manufacturers, which means costs are elevated — even before considering logistics and labor.
As mentioned by Millionaire MNL, tech giants like Meta and Microsoft have begun lobbying quietly for tariff exemptions tied to AI infrastructure, arguing that compute should be treated as a national asset.
But for now, the tariffs remain — adding complexity to an already expensive AI buildout.
Wall Street still backing the plan
Despite the eye-watering numbers, investors appear to be holding the line. Meta’s stock has remained strong, with analysts viewing the spending surge as strategic rather than reckless. The company’s high-margin ad business continues to deliver profits, giving Zuckerberg room to spend big while maintaining financial credibility.
Still, the scale of Meta capex inflation raises questions about sustainability — and whether other firms can keep pace in the AI arms race without similar war chests.
If Meta hits the upper end of its $72 billion forecast, it would surpass the total GDP of several small countries — just to build infrastructure for the next wave of digital intelligence.