#美股财报季 #Gate广场万圣节 #每周来晒 The “soul-searching question” has finally arrived: Can it win the AI “cash-burning” race?

On October 22, Beijing time, Meta released its third-quarter 2026 earnings report, with revenue up 26% year over year to $64.601 billion, capital expenditures of $34.91 billion for the quarter, and free cash flow turning negative. As revenue grows alongside expanding investment, AI monetization is facing a more concrete test: Can paid demand translate into sustained profits, and can cash generated by operations keep pace with investment in computing power?
This quarter, advertising remained the main driver of Meta’s revenue growth. Advertising revenue was $63.061 billion, up 26% year over year; ad impressions increased 10%, while the average ad price rose 14%. The company said AI is generating incremental ad revenue by improving recommendations, user engagement, and monetization efficiency.
Alongside business expansion, a steadily growing bill is also coming due. Meta revised its full-year capital expenditure guidance from $130 billion to $145 billion to $135 billion to $145 billion.
Chief Financial Officer Susan Li said Meta’s capital expenditures would continue to increase over the next few years, mainly to support its core business and emerging AI initiatives. Its personal agent Muse is a new AI product carrying high expectations. While the company has not separately disclosed Muse’s revenue or profit, Zuckerberg said on the earnings call that more than 10 million users in the United States and Canada had used Muse within about a month and a half of its launch. He also acknowledged that Muse costs more to run than many of the company’s previous new products, and that Meta would push monetization earlier than before. He said the next three priorities are scaling the product, increasing revenue, and improving computing costs, and he expects Muse’s computing cost efficiency to improve more than tenfold over the next year. Progress in AI is also shaping external expectations. Morgan Stanley raised its earnings forecasts for Meta in 2027 and 2028, citing AI-driven ad growth and Muse’s revenue potential.
Similarly, Anthropic, developer of the Claude model, is showing a comparable trend. Earlier reports disclosed that Anthropic posted a net loss of $80.7 billion in the first half of the year, including $72 billion in non-cash accounting expenses from changes in the fair value of convertible bonds. In the third quarter, Anthropic generated revenue of $16.8 billion, up about 45% quarter over quarter, and achieved adjusted profitability for the quarter. Anthropic reportedly invested nearly $9 billion in computing power and infrastructure in the first half of the year, exceeding its comparable full-year 2025 spending of about $7.3 billion; it has also committed to investing at least $518 billion in cloud services, computing, and infrastructure over the next ten years.

The AI “cash-burning” race is intensifying. Whether it is Meta, with its enormous advertising revenue, or Anthropic, which is about to pursue one of the largest IPOs in history, both face a shared challenge: The capital being committed is premised on AI companies’ ability to generate sufficient revenue in the future. Training more powerful models and expanding the user base both require computing capacity to be secured in advance. How to seize market opportunities while balancing financial pressure and investors’ return expectations is a question AI companies will need to keep answering.