Meta Platforms faces uncertainty over AI return on invested capital as capex balloons to $145B

Meta Platforms just told investors it plans to spend between $125 billion and $145 billion on capital expenditures in 2026. Most of that money is flowing into AI infrastructure, including data centers and GPUs. The number is staggering on its own, but it lands differently when you remember what happened the last time Meta went all-in on a big technology bet.

The company’s Q1 2026 earnings report showed revenue of $56 billion, a 33% jump year-over-year. Morningstar has slapped a High Uncertainty Rating on Meta, and the core question haunting the stock is straightforward: will any of this spending actually generate returns that justify the price tag?

The metaverse ghost still lingers

Meta’s cumulative metaverse spending has surpassed $83 billion since 2022. The returns from that initiative have been minimal. Reality Labs continues to operate as a money furnace, and investors who watched those billions evaporate have understandable trust issues when the company asks them to believe in the next massive capital deployment cycle.

Mark Zuckerberg has described conversations about return on investment as “very technical.” His framing emphasizes a long-term commitment to building advanced AI models that could cement Meta’s dominance.

The difference this time is that AI already has a visible impact on Meta’s core advertising business. The metaverse never had that advantage.

Where the AI money is actually working

AI-driven improvements to ad targeting and user engagement have been identified as key growth factors behind that 33% revenue increase. Some internal metrics suggest the return on invested capital from AI-enhanced advertising exceeds 20%.

Free cash flow is where the math gets uncomfortable. Analysts project that Meta’s free cash flow will come under increasing pressure as capital expenditures surge, and some projections suggest it could turn negative.

The broader hyperscaler arms race

Meta isn’t alone in this spending spree. The broader trend among hyperscaler companies involves collective plans to invest hundreds of billions annually in AI infrastructure. Microsoft, Google, and Amazon are all running similar playbooks, pouring capital into data centers, custom chips, and AI model training at unprecedented scale.

Analysts predict that clarity regarding whether these AI investments will deliver meaningful returns should emerge over the next two earnings cycles.

What this means for investors and crypto-adjacent markets

The crypto angle here is more indirect but still worth noting. Meta’s previous foray into digital assets through the Libra/Diem stablecoin project ended in failure after regulatory pushback. The company currently maintains restrictive policies on crypto advertising across its platforms.

GPU demand from hyperscalers has already influenced the economics of crypto mining operations, as competition for high-end chips drives up costs.

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