Meta Platforms is spending up to $145 billion on artificial intelligence infrastructure in 2026, nearly double what it spent in 2025, but the company's leadership has offered few concrete details about when or how this massive investment will pay off. During Meta's first quarter earnings call in late April, CEO Mark Zuckerberg deflected questions about return on investment (ROI), calling it "a very technical question," while the company's stock tumbled more than 6% in after-hours trading following the announcement of the increased spending guidance.

The jump in capital expenditure guidance caught investors off guard. Meta raised its full-year 2026 capex forecast to a range of $125 billion to $145 billion, up from a previous estimate of $115 billion to $135 billion. The company attributed the increase to higher component prices and additional data center costs needed to support future-year capacity. To put this in perspective, Meta spent $72.2 billion on capital expenditures in 2025, meaning the company is now planning to spend nearly double that amount in 2026.

Why Is Meta Spending So Much on AI Infrastructure?

Meta's massive infrastructure investment is driven by its ambition to build what it calls a "leading lab in the world" for artificial intelligence. The company views AI as essential to its core advertising business and longer-term investments in personal AI agents for business, health, and entrepreneurship. Zuckerberg has explained that these investments will strengthen the advertising business by making recommendations more relevant and improving how ads are targeted to increase the time consumers spend on Meta's platforms, including Instagram, WhatsApp, and Facebook.

During the earnings call, Zuckerberg outlined a vision for how Meta's new AI models will transform the company's approach to content recommendation. "For the first time in Meta's history, we're going to be able to develop a first-principles understanding of what you care about and what each piece of content in our system is about," he explained. "So that way, we can show you more useful things for what you're trying to accomplish and we'll also be able to create personalized content specifically for people to help you achieve your goals as well".

Zuckerberg

What Is Meta Actually Building With This Money?

Meta is constructing a diversified hardware strategy to support its AI ambitions. The company is rolling out more than one gigawatt of custom silicon chips developed in partnership with Broadcom, along with significant quantities of AMD chips to complement new Nvidia systems. This multi-vendor approach suggests Meta is trying to reduce its dependence on any single chip supplier and manage costs across its sprawling data center infrastructure.

On the software side, Meta's Superintelligence AI lab has released what Zuckerberg described as a "significantly upgraded" version of Meta AI. The company claims to have built "the strongest research team in the industry" over the past 10 months and established the scientific and technical foundations to scale very advanced models. However, Zuckerberg was notably vague about specific milestones or timelines for demonstrating returns on the infrastructure investment.

How to Evaluate Meta's AI Investment Strategy

• Financial Performance Metrics: Watch for improvements in advertising efficiency and click-through rates on Meta's platforms. The company claims AI will make recommendations more relevant, which should theoretically increase user engagement and advertising revenue per user.
• Model Capability Benchmarks: Monitor releases of Meta's AI models and how they compare to competitors like OpenAI's GPT series and Anthropic's Claude on industry-standard performance tests. Zuckerberg mentioned building "leading models," so performance data will be crucial.
• Product Adoption Rates: Track how many users actually engage with new AI-powered features across Facebook, Instagram, and WhatsApp. Zuckerberg's strategy depends on getting AI features "to billions of people," so adoption metrics will signal whether the investment is working.
• Cost Efficiency Improvements: Assess whether Meta's custom silicon and multi-vendor chip strategy actually reduce per-unit computing costs over time. This will determine whether the company can sustain such high spending levels.

The investor skepticism is understandable. Melissa Otto, head of Visible Alpha Research at S&P Global, noted that while Meta had a solid quarter with first quarter revenues of $56.3 billion (up 33% year-over-year) and operating income of $22.9 billion (up 30%), the company's guidance raised questions about what the real return on all this capital spending will be. "I think the investment community is getting a little frustrated at the amount of cash they're burning," Otto observed.

Interestingly, Meta's earnings performance contrasted with how investors reacted to similar announcements from competitors. Alphabet and Amazon, which are also spending enormous sums on AI infrastructure, both saw their share prices rise after hours on the same day Meta's stock fell. Both companies reported AI-related growth in their massive cloud services businesses, giving investors concrete evidence that their AI spending was generating returns.

Zuckerberg's response to questions about ROI suggests the company doesn't yet have a precise formula for measuring success. When asked about signposts or key factors Meta is watching to ensure it's "on the right path" to generating healthy returns over the next 12 to 24 months, Zuckerberg said the company is focused on "building leading models and leading products" and getting them "to billions of people" before worrying about monetization. "I think we have a sense of the shape of where these things need to be," he added, but acknowledged the company doesn't have "a very precise plan for exactly how each product is going to scale month over month".

Zuckerberg

The company is also making significant personnel changes to support its AI ambitions. Meta's chief financial officer Susan Li noted that the growth in employee compensation was driven by technical hires added over the past year, particularly AI talent. However, the company also announced plans to "reduce the size" of Meta's employee base in May, with reports suggesting hundreds of jobs would be cut in the United States and abroad among teams including sales, recruiting, and hardware.

For now, Meta is betting that its massive infrastructure investment will eventually translate into competitive advantages in AI that strengthen its core advertising business and unlock new revenue streams from AI-powered products and services. But without clearer milestones or more specific ROI metrics, investors will likely remain skeptical about whether the company's $145 billion spending plan represents a smart long-term investment or an expensive gamble on uncertain returns.