Meta Platforms is preparing to spend as much as $145 billion on artificial intelligence infrastructure by 2026, a figure that dwarfs its previous capital expenditure plans and signals an aggressive bet on AI-driven growth. The company’s projected outlay, disclosed in internal planning documents, covers data centers, specialized chips, and energy systems needed to train and run large-scale AI models. Morningstar, the investment research firm, has assigned a “High Uncertainty Rating” to the returns Meta can expect from that spending, citing the unpredictable nature of AI monetization and competitive pressures.
What the $145 billion covers
The capital expenditure plan spans the next three fiscal years, with the bulk expected in 2026. Meta has not publicly confirmed the exact figure, but the documents reviewed by GFdaily show the company is budgeting for a massive expansion of its computing capacity. That includes building new data centers, purchasing Nvidia graphics processing units, and developing its own custom AI accelerator chips. The spending also covers energy infrastructure, as AI workloads require enormous amounts of electricity. Meta’s previous record capex was $39.3 billion in 2024, making the 2026 target roughly 3.7 times larger.
Morningstar’s uncertainty rating
Morningstar analysts assigned the High Uncertainty Rating after evaluating Meta’s AI investment against potential revenue streams. The firm noted that while Meta has a strong track record in advertising and social platforms, the path to monetizing AI at scale remains unclear. Competitors like Google, Microsoft, and Amazon are also pouring billions into AI, raising the risk that returns could be diluted by a race to the bottom on pricing or by rapid technological shifts. Morningstar’s rating is its second-highest uncertainty level, reserved for companies where the range of possible outcomes is wide and the margin of safety is thin.
Why Meta is betting so big
Meta CEO Mark Zuckerberg has framed AI as the company’s top priority, arguing that advanced AI systems will unlock new products, improve advertising targeting, and eventually power the metaverse. The company’s AI research division, FAIR, has released open-source models like Llama, and Meta is integrating AI into its core apps—Facebook, Instagram, and WhatsApp. The spending plan suggests Meta believes it must own its infrastructure to stay competitive, rather than relying solely on cloud partners. However, the sheer scale of the investment has raised eyebrows among some investors, who worry about a repeat of Meta’s earlier overinvestment in the metaverse, which cost the company billions with limited returns.
What happens next
Meta is expected to provide more details on its capex plans during its next earnings call, scheduled for late April. Investors will be watching for signs of how the company plans to generate revenue from its AI spending, including potential new products or advertising tools. Meanwhile, Morningstar’s uncertainty rating may influence how some funds allocate to Meta stock. The question hanging over the company is whether its AI bet will pay off before the spending becomes a drag on earnings.
