CBN News, October 27, 2026, 12:20:11 - Listen to the news
Author: Liu Xiaojie Editor-in-Chief: Li Na
As of the end of September, Meta still had approximately $82 billion in cash and marketable securities on its account. However, under the pressure of computing power costs, the company is still preparing to seek a large sum of money from the capital market.
On October 26th local time, Meta announced a financing plan totaling approximately $95 billion to fund the infrastructure of AI and the expansion of "personal super intelligence."
As mentioned in the announcement, Meta, this financing plan is divided into two parts. One part consists of a simultaneous underwritten public offering of approximately $35 billion, which includes about $22.5 billion in Class A common shares and about $12.5 billion in depositary receipts representing preferred stocks. In addition, the company will launch a plan ( ATM ) with a scale of approximately $60 billion to be issued at market price.
The so-called issuance at market price means that the company will sell shares in batches according to market conditions. The actual issuance time, scale, and terms have not yet been determined, and the company expects it to be no earlier than the first quarter of 2027.
Based on the previous estimate of $135 billion to $145 billion in capital expenditures for 2026 by Meta, $95 billion exceeds two-thirds of that forecasted capital expenditure.
The company stated that NVIDIA, AMD, and Broadcom have expressed their intention to purchase up to $5 billion in Class A common shares as cornerstone investors, but these intentions do not constitute binding subscription commitments. The AI infrastructure competition continues, and the linkage between capital and computing power is becoming increasingly strong.
Regarding the reasons for financing, Meta mentioned that they are investing in expanding infrastructure and computing power to support the opportunities brought by AI, with the aim of laying out personal super intelligence.
In order to make the financing more persuasive, Meta listed the company's current progress in the field of AI. The most notable achievement is the personal AI intelligent entity Muse.
According to Meta, within a few weeks of its launch, Muse attracted over 10 million users in the United States and Canada. Shopping is its main use case, with more than 25% of weekly active users participating in shopping activities. Mark Zuckerberg of Meta CEO stated during a previous financial report conference call that the average purchase amount per user exceeded $200, which included travel expenses.
However, the supply of computing power has constrained the promotion of this AI product. Zuckerberg stated that the company is building computing capabilities that will be able to serve over a billion people in the coming years, but Muse is more compute-intensive, and efficiency and capacity are important limiting factors for further expanding coverage.
But the problem is that the more users there are, the larger the bills may also be. While building computing power, Meta has also proposed to increase the computational cost efficiency of Muse by more than ten times in the coming year, by making more efficient use of computing power during off-peak hours, adopting more efficient models, and optimizing inference processes, thereby reducing the cost per service.
Currently, Meta has data centers in 23 states of the United States and is building four super-clusters with a capacity of over 1 gigawatt each, with the potential to nearly double the total infrastructure capacity by 2026.
For Meta, product expansion, revenue growth, and cost optimization must be pursued simultaneously. Zuckerberg stated that Muse will start generating revenue earlier than the company's previous products, due to both operating costs and the commercial needs that users have already demonstrated. Management expects that by expanding the API, Muse Code, and enterprise agent businesses, the new AI business is expected to contribute to incremental revenue starting from 2027.
From a financial data perspective, Meta is still within the growth range. The third-quarter report released on October 21st exceeded market expectations: revenue for the third quarter of 2026 was 64.6 billion US dollars, a year-on-year increase of 26%; revenue for the fourth quarter is estimated to be between 73 and 76 billion US dollars.
On the other hand, negative free cash flow has become a focal point of market attention. In the third quarter, Meta's capital expenditures, including the repayment of financing lease principal, amounted to $34.91 billion, exceeding the $32.79 billion in operating cash flow, resulting in a negative free cash flow of $2.12 billion.
Meta is not the only company to experience negative quarterly free cash flow and that also needs to raise a substantial amount of funds for the expansion of AI.
In June this year, Google's parent company announced the largest financing in its history: it plans to raise $85 billion to invest in the infrastructure construction of AI. In the financial report for the second quarter of 2026, Google reported a negative free cash flow of $5.9 billion for the first time in its history.
What both companies have in common is that they are still able to generate substantial cash from their existing operations, but they aim to build a larger scale of AI infrastructure with more concentrated investment, while the revenue from new businesses will take time to grow.
Recently, AI unicorn Anthropic is also making preparations for IPO. The prospectus reveals that in the first half of Anthropic2026, the company generated revenue of $16.29 billion, but incurred a net loss of $80.75 billion. Anthropic plans to commit at least $518 billion in expenditures over the next decade in areas such as cloud services, computing power, and infrastructure.
Tech giants are increasingly turning to the capital market, showing a greater reliance on external financing, which reflects that the funding needs of the entire AI industry continue to grow. Goldman Sachs estimates that from 2026 to 2031, the total capital expenditures related to AI could reach $7.6 trillion, including costs for computing power, data centers, and power infrastructure.
The successfully funded AI giants may be able to significantly expand their infrastructure, but whether they can convert this into revenue and profit in the future and cover the costs of computing power remains unknown for investors. As of the close on October 26th local time, Meta stocks fell by 1.88% to $630.56 per share, with a current total market value of 1.61 trillion dollars.
