Meta Plans $95 Billion Capital Raise as AI Infrastructure Costs Escalate

Meta Platforms plans to raise up to $95 billion through equity offerings to finance its artificial intelligence infrastructure and superintelligence ambitions, highlighting how the race to develop advanced AI systems is pushing even highly profitable technology companies to seek additional capital.

The company said on Monday that it would initially raise $35 billion through underwritten public equity offerings, followed by another $60 billion through an at-the-market programme that allows shares to be sold gradually into the market. Nvidia, Advanced Micro Devices and Broadcom have expressed interest in purchasing up to $5 billion in shares, Meta said.

Meta shares fell 1.6% in morning trading following the announcement, as investors weighed the fundraising against the company's rising capital expenditure and concerns about its ability to generate cash after funding its infrastructure expansion.

The proposed raise comes amid an unprecedented investment cycle across the technology industry. Goldman Sachs estimates that AI-related capital expenditure could reach approximately $7.6 trillion between 2026 and 2031, spanning computing infrastructure, data centers and power generation.

Meta said the proceeds would support emerging AI opportunities across its family of applications, its Muse AI agent and its data center operations. The company plans to finance these investments through a combination of operating cash flow, investment-grade debt, partnerships and equity capital.

The decision to seek substantial equity financing suggests Meta is broadening its funding options as the cost of developing and operating advanced AI systems increases. Although the company continues to generate significant revenue from its advertising business, its infrastructure ambitions require capital commitments that could absorb a growing share of internally generated cash.

Equity financing also offers a different trade-off from borrowing. Selling shares does not create interest payments or require repayment of principal, but it dilutes existing shareholders' ownership. The investment will ultimately need to generate sufficient additional earnings and cash flow to justify that dilution.

Meta's fundraising plans follow a third-quarter performance in which revenue growth exceeded expectations but failed to prevent concerns about the financial consequences of its AI spending.

The company reported revenue of $64.60 billion, a 26% increase from a year earlier and above analysts' expectations of $63.37 billion. However, investors focused on negative free cash flow and a higher full-year capital expenditure outlook.

Meta now expects capital expenditure of between $135 billion and $145 billion for the year, raising the lower end of its previous forecast from $130 billion. The revised outlook contributed to a decline in its shares on Thursday.

Negative free cash flow is not necessarily evidence of deteriorating business performance when a company is investing heavily in long-lived assets. However, it indicates that capital expenditure is exceeding the cash available after operating requirements, leaving less internally generated funding for additional projects, acquisitions or shareholder returns.

But that matters as Meta expands its computing capacity. The company is building four superclusters, each with more than one gigawatt of capacity, and expects to nearly double its total infrastructure capacity in 2026 compared with the previous year.

These facilities require extensive investment in computing equipment, networking, cooling systems, data-center construction, and electricity supply. The spending often precedes the revenue that the infrastructure is expected to generate, creating a gap between the timing of capital commitments and the commercial returns.

Meta's challenge is to convert this capacity into measurable economic benefits. AI could improve advertising targeting, increase user engagement, support new paid services and create additional business opportunities. Its Muse AI agent and other AI products could also help the company expand beyond its traditional advertising model.

But analysts say the financial payoff will be determined by adoption, pricing, and operating efficiency. Greater computing capacity does not automatically translate into proportionate revenue growth, especially if competitors offer similar services or if the cost of operating advanced models remains high.

Ben Wood, senior partner at CCS Insight, said the decision to raise more capital "reflects the eye-watering levels of investment required to keep up in the AI arms race."

"Meta offering more stock comes at a time when numerous companies, such as Google, SpaceX and soon, Anthropic, are going to the markets to raise funds," he said.

Technology Giants Turn to Capital Markets

Meta's proposed $95 billion raise illustrates a wider shift in how major technology companies are financing the AI infrastructure build-out. Large technology businesses have historically relied heavily on operating cash flow, but the scale of current investment is encouraging a broader mix of equity, debt and infrastructure partnerships.

Alphabet, Google's parent company, said in June that it planned to raise $85 billion. SpaceX has closed a $40 billion debt financing to purchase graphics processing units, while Anthropic is preparing for an initial public offering.

TikTok US is also seeking to secure multiple gigawatts of computing capacity over the coming years, CNBC reported on Friday, adding to anticipated demand for the infrastructure needed to develop and deploy AI systems.

The parallel fundraising efforts raise questions about the sustainability of the investment cycle. Goldman Sachs' $7.6 trillion estimate points to the scale of the resources that could be committed to AI over the next several years, but the amount spent will not, by itself, determine whether those investments generate adequate returns.

The economics are tied to how efficiently companies use their computing capacity, the availability and cost of electricity, the pace of AI adoption, and their ability to turn technical advances into products customers will pay for. Strong demand could allow infrastructure owners to build durable competitive advantages. Slower monetization, however, could leave companies carrying expensive assets that generate returns below expectations.