Category: AI News

The autumn rush of artificial intelligence companies to the stock market has narrowed to a single name. Anthropic is pressing ahead with plans to hold its initial public offering in mid-November, even as a string of other AI-related listings have been shelved or postponed, according to people familiar with the matter.

The company’s prospectus could become public as early as late October, and Anthropic is expected to begin marketing its shares to prospective investors during the week of November 9, Reuters reported on Thursday. The people cautioned that the timetable remains subject to change.

It is an unusually lonely path to the public markets. A technology IPO calendar that looked busy a few months ago has been thinned by market jitters: Switch, Nscale, SB Energy, Elice, AirTrunk REIT and Firmus have all delayed or withdrawn their listings. Investors have grown more cautious as interest rates rise and shares of AI infrastructure operators such as CoreWeave sell off, prompting harder questions about whether the industry’s enormous data-centre spending will ever be paid back.

Nscale’s delay was partly driven by comments from staff at the US Securities and Exchange Commission, underscoring the regulatory glare now falling on the sector.

The numbers inside Anthropic’s draft prospectus, reviewed by The Wall Street Journal, help explain both the confidence and the questions. Revenue reached $16.3 billion in the first half of 2026, but the company recorded a net loss of more than $80 billion over the same period — largely an accounting adjustment tied to financing agreements, rather than day-to-day cash losses. Roughly one-fifth of that revenue came from just two customers, a concentration that will draw scrutiny from new investors.

The document also signals that Anthropic could spend as much as $750 billion in the coming years on computing capacity, including agreements that are still being negotiated — a figure that shows just how far the industry’s appetite for chips and data centres still stretches.

On valuation, expectations remain stratospheric. People familiar with the company’s thinking believe a figure near $1.5 trillion is plausible, while the Journal’s review of the draft prospectus referenced a potential $2 trillion valuation.

Anthropic did not immediately respond to Reuters’ request for comment.

Sources: Reuters (https://www.reuters.com/legal/transactional/anthropic-ipo-stands-alone-ai-market-jitters-sideline-others-2026-10-22/) and The Wall Street Journal (https://www.wsj.com/tech/ai/anthropic-revenue-growth-ipo-accaf0f1).

DeepSeek, the Chinese artificial intelligence startup that startled Silicon Valley last year by building a frontier-class model at a fraction of the expected cost, has raised $15 billion in its second private funding round — nearly double its original target — as it accelerates plans to list in Shanghai early next year.

The round, which ran from July through this month, drew in both existing shareholders and new heavyweight backers, according to people familiar with the matter. Battery giant Contemporary Amperex Technology (CATL) and tech group Tencent, both previous investors, put in more money, alongside existing shareholders Monolith Management and Shixiang. New participants include semiconductor-focused private equity firm Stony Creek Capital, Legend Capital and carmaker Geely. Before the round closed, DeepSeek was valued at roughly $74 billion.

The company is now moving quickly toward public markets. It has hired state-owned investment bank Citic Securities to advise on the listing and recruited former venture capitalist Yan Wentao as chief financial officer in September. If the timetable holds, DeepSeek could reach the stock market ahead of OpenAI — a striking prospect for a lab that, less than two years ago, was known mainly to researchers.

DeepSeek’s appeal to investors rests on an unusual combination: competitive models and, by the standards of the industry, striking efficiency. Founder Liang Wenfeng told investors in September that the company’s annualised revenue run rate had reached $1 billion, more than double the $400 million to $500 million reported in July. The money comes mainly from cloud-based access to its models through APIs. The company has reportedly maintained gross margins of 70% to 80% on access to its V4 flagship model despite charging less than US rivals, and in August it introduced peak-hour pricing that raised rates for its flagship models roughly fourfold at busy times — still below major competitors’ prices.

The raise also lands at a moment when investors globally are asking harder questions about AI economics. On the same day DeepSeek’s round was reported, US tech shares were sliding precisely because of anxiety over whether heavy AI infrastructure spending will generate adequate returns. DeepSeek’s pitch to investors is, in effect, the opposite one: comparable capability without the comparable bill. Annualised revenue is not profit, however, and the company continues to invest heavily in computing power — the new capital is expected to fund model research and infrastructure as it prepares its restructuring for the listing.

The deal is the largest primary-market financing in China this year and underlines how much of the AI race is now being funded and fought outside the United States. Rival Chinese lab Moonshot AI is also reported to be approaching a stock market debut. For Washington, which has tried to slow China’s AI progress by restricting exports of advanced chips, DeepSeek’s trajectory — efficient models, a $74 billion pre-IPO valuation and a domestic listing — is an uncomfortable counterpoint.

Sources:
The Wall Street Journal via Morningstar — https://www.morningstar.com/news/dow-jones/202610223804/chinese-ai-startup-deekseek-raised-15-billion-ahead-of-potential-ipo
PYMNTS — https://www.pymnts.com/news/artificial-intelligence/2026/deepseek-raises-15-billion-ahead-of-planned-shanghai-ipo