Meta delivered strong fourth-quarter earnings on Wednesday, beating Wall Street expectations with $59.89 billion in revenue versus the estimated $58.41 billion, and earnings per share of $8.88 compared to expectations of $8.19. The stock surged over 8% in after-hours trading following the announcement.
The headline story, however, is Meta’s massive AI infrastructure investment. The company announced plans to spend between $115 billion and $135 billion on capital expenditures in 2026, significantly exceeding analyst estimates of $110.62 billion. Meta’s actual 2025 capex came in at $72.2 billion, with $22.1 billion spent in Q4 alone, already surpassing expectations as the company continues its aggressive AI buildout.
CEO Mark Zuckerberg made a forceful case for Meta’s AI independence during the analyst call, positioning Meta as a “deep technology company” that must develop its own frontier AI models rather than relying on third-party APIs. He argued that frontier models may not always be available via API in the future, making it critical for Meta to maintain its own capabilities. This puts Meta in direct competition with OpenAI, Anthropic, and xAI in the crowded large language model space.
Despite the AI push, advertising remains Meta’s core business. When pressed by Bank of America analyst Justin Post about whether Meta can “do things beyond ads,” Zuckerberg acknowledged that ads will continue to be the company’s “driver of growth” for the next couple of years. However, he outlined how AI will enhance the advertising experience through improved recommendations and LLM integration.
Meta is already seeing AI impact across its business. CFO Susan Li reported that the combined revenue run rate of Meta’s AI-powered video generation tool hit $10 billion in Q4. The company is also rolling out AI business assistants for advertisers and has AI-translated videos generating hundreds of millions of views across nine languages.
Capacity constraints remain a challenge despite the massive spending. Li stated that Meta expects to remain capacity-constrained through much of 2026 until additional data center facilities come online later in the year. The company is even exploring leasing out data center space while simultaneously claiming it needs more capacity.
Zuckerberg also teased future AI applications, including “agentic shopping” tools that will help users find specific products, and AI-powered game creation in Horizon Worlds. He noted that Meta is already seeing “projects that used to require big teams now be accomplished by a single, very talented person” thanks to AI tooling, suggesting potential workforce implications ahead.
Key Quotes
We’re a deep technology company — more than just the apps for which it’s known. I think that frontier models will not always be available via an API in the future. That means it’s important for Meta to have its own frontier model.
Mark Zuckerberg explained Meta’s rationale for developing its own large language models rather than relying on third-party providers. This statement signals Meta’s concern about future AI infrastructure dependencies and justifies the massive capital expenditure on proprietary AI development.
For the next couple of years, ads are going to be our driver of growth.
Zuckerberg’s frank admission to Bank of America analyst Justin Post acknowledges that despite massive AI investments, Meta’s traditional advertising business will remain its primary revenue source in the near term. This highlights the gap between AI spending and AI monetization that Meta and other tech giants currently face.
We’re already seeing projects that used to require big teams now be accomplished by a single, very talented person.
Zuckerberg described how AI tools are transforming productivity within Meta itself, suggesting that AI-powered automation could enable the company to accomplish more with fewer employees. This has significant implications for the future of work and workforce composition in tech companies.
We expect over the course of 2026 to have significantly more capacity this year as we add cloud, but we’ll likely still be constrained through much of 2026 until additional capacity from our own facilities comes online later in the year.
CFO Susan Li explained Meta’s ongoing capacity constraints despite massive infrastructure spending. This reveals the enormous computational demands of frontier AI development and suggests that even unprecedented capital investment struggles to keep pace with AI’s resource requirements.
Our Take
Meta’s earnings call reveals a company making an all-in bet on AI supremacy, even as it grapples with the fundamental challenge facing Big Tech: how to monetize these massive investments. The $115-135 billion capex guidance is staggering, but what’s more telling is Zuckerberg’s defensive posture about needing proprietary AI models. This suggests anxiety about being locked out of critical AI infrastructure by competitors.
The most significant revelation may be the productivity claims—that single engineers can now accomplish what once required entire teams. If true across the industry, this portends major workforce restructuring ahead. Yet Meta’s 6% headcount growth, concentrated in technical roles, suggests the company is simultaneously hiring aggressively for AI talent while using AI to reduce headcount needs elsewhere.
The gap between AI spending and AI revenue generation remains Meta’s biggest vulnerability. While the $10 billion revenue run rate for AI video tools is impressive, it’s dwarfed by the infrastructure costs. Meta is essentially asking investors to trust that today’s spending will unlock tomorrow’s revenue streams—a bet that could take years to validate.
Why This Matters
Meta’s earnings reveal the staggering financial commitment required to compete in the AI race. The $115-135 billion capex guidance for 2026 represents one of the largest infrastructure investments in tech history, signaling that the AI arms race is entering an unprecedented spending phase. This has major implications for the entire industry, as competitors like Google, Microsoft, and Amazon face pressure to match these investments or risk falling behind.
Zuckerberg’s emphasis on AI independence highlights growing concerns about API dependency. His argument that frontier models may not always be available suggests fears about supply chain vulnerabilities or potential monopolization in the AI infrastructure layer. This could accelerate vertical integration across the tech industry.
The tension between AI investment and current revenue models is becoming acute. While Meta pours over $100 billion into AI infrastructure, it acknowledges that advertising will remain its primary revenue driver for years. This raises questions about ROI timelines and whether AI investments will generate proportional returns. For businesses and investors, Meta’s approach represents a high-stakes bet that AI capabilities will eventually unlock new revenue streams beyond traditional advertising, fundamentally reshaping how social media platforms monetize their massive user bases.
Related Stories
- Meta and Nvidia Billionaires’ Wealth Soars $152B in AI Boom
- Big Tech’s 2025 AI Plans: Meta, Apple, Tesla, Google Unveil Roadmap
- What Zuckerberg Risks by Following Musk’s Lead
- Groq Investor Warns of Data Center Crisis Threatening AI Industry
- CEOs Express Insecurity About AI Strategy and Implementation
Source: https://www.businessinsider.com/meta-q4-earnings-stock-price-ai-capex-live-updates-2026-1