Inside Meta's US$135bn AI Infrastructure Buildout

Meta Platforms has entered a new phase of capital-intensive growth, with financial results for 2025 revealing both the sustained profitability of its advertising operations and the mounting costs of its artificial intelligence ambitions.

The social media giant's fourth quarter performance demonstrated the financial strength underpinning its strategic pivot towards AI infrastructure.

Revenue reached US$59.9bn in the three months ending December, marking a 24% year-on-year increase that surpassed analyst forecasts.

Annual revenue climbed 22% to US$201bn, while quarterly net income rose 9% to US$22.8bn.

However, the results also exposed the financial pressures emerging from Meta's accelerated investment programme, with operating margins compressing as the company channels resources into technical talent and infrastructure.

Investment spending reaches unprecedented levels

Meta's capital expenditure plans for 2026 represent a dramatic escalation in financial commitment to AI development. The company projects spending between US$115bn and US$135bn this year, nearly double the US$72.2bn deployed in 2025.

Chief Financial Officer Susan Li indicated during the 28 January 2025 earnings call that the majority of this capital will fund data centres, servers and networking equipment necessary for training advanced AI models.

This investment trajectory means Meta has now allocated approximately US$140bn to AI-related initiatives over a three-year period.

Mark Zuckerberg, the company's Founder and CEO, framed the spending as essential to competitive positioning. "We are now seeing a major AI acceleration," he explained, adding that Meta would "continue to invest very significantly in infrastructure to train leading models and deliver personal superintelligence to billions of people and businesses around the world".

The financial markets responded positively to the results despite the increased expenditure, with Meta's share price rising between 6% and 10% in extended trading following the announcement.

Operating margins face compression pressure

While revenue growth remained robust, profitability metrics showed signs of strain. The operating margin contracted to 41% in the fourth quarter, down from 48% in the same period of 2024, as costs and expenses surged 40% to US$35.1bn.

Susan attributed the cost increases to deliberate strategic choices, noting that "year-over-year growth was driven primarily by employee compensation expenses, legal expenses and infrastructure costs".

She specified that higher compensation reflected "the technical hires we've added this year, particularly AI talent".

This margin compression raises questions about Meta's ability to sustain profitability levels whilst simultaneously funding its AI transformation.

However, Susan expressed confidence in the company's financial trajectory, stating that "in 2026 we expect to deliver operating income that is above 2025 operating income" despite the infrastructure investment increase.

The company's balance sheet appears positioned to support this expansion, with US$81.6bn in cash and marketable securities at year-end 2025. Meta anticipates funding its growth primarily through operating cash flow rather than external financing.

Advertising revenue growth masks structural shifts

Meta's advertising business continued to generate substantial cash flow, with ad revenue of US$58.1bn in the fourth quarter representing 24% growth compared with the prior year.

The increase stemmed from an 18% rise in ad impressions combined with a 6% increase in average price per advertisement.

The company's user base expansion provided the foundation for advertising growth, with family daily active people averaging 3.58bn in December, up 7% annually. Facebook, Instagram and WhatsApp each recorded approximately two billion daily users or more.

Yet beneath these figures, Meta is experiencing operational transformation driven by AI-enabled productivity gains. Susan noted that "since the beginning of 2025, we've seen a 30% increase in output per engineer" due to AI-powered development tools.

Mark suggested these efficiency improvements reshape workforce requirements, observing that "we're starting to see projects that used to require big teams now be accomplished by a single very talented person".

This comment followed recent redundancies, including several hundred positions eliminated from the Reality Labs division focused on virtual reality and metaverse development.

The financial results illuminate Meta's strategic calculation: leveraging advertising profitability to fund a capital-intensive transition towards AI leadership, whilst accepting near-term margin pressure in pursuit of longer-term competitive advantage.