Artificial intelligence developer Anthropic took the first step Monday toward becoming a publicly traded company, a move that will give it access to a large pool of investor capital when it goes public.
Anthropic said Monday in an announcement that it has privately filed a draft S-1 with the US Securities and Exchange Commission, allowing the company to go public after an SEC review. Anthropic said it has not set the number of shares to be offered or any prices, and that the move “will depend on market conditions and other factors.”
An Anthropic representative did not immediately respond to questions about the expected timing or amount of the offering.
Claude Maker is one of three major tech companies expected to have initial public offerings this year amid what some are calling “AI gold.” SpaceX, the rocket company owned by Elon Musk that also includes the Starlink ISP, the AI lab xAI, and the social network now known as iX, filed for IPO in May. Anthropic’s main competitor, the maker of ChatGPT is OpenAI expected to follow soon.
The frenzied IPO race reflects the market’s eagerness to cash in on its multibillion-dollar bet, as AI companies rush to secure the big funding needed to survive. The AI industry is capital intensive, driven by the huge costs of maintaining the computing power needed to train large linguistic models, as well as the data centers, silicon and power grids to keep them running.
Although Anthropic has turned in voluntary documents for legal review, that does not guarantee a final decision on the IPO, and the company may decide not to go public, according to Patrick Corrigan, a law professor at the University of Notre Dame. Based on the SEC’s usual timelines, a public filing can be expected in a few weeks, with stock trading likely to begin in two to four months, Corrigan told CNET in an interview.
Income volatility?
The AI industry has become a highly speculative environment, where valuation is determined by a company’s future potential rather than current profitability. An online tracker of revenue and losses found that twice as much money was spent on AI development as it was returned, pointing to billions of dollars in debt. The only major company to come out on top is Nvidia, which makes gold-core AI chips.
Critics point out that AI companies have raised money by using deceptive accounting, using “annual” revenue spikes and ignoring core costs to hide poor margins, thereby misleading investors.
“Right now, their valuations are so high that it’s becoming increasingly impossible to raise a lot of money, and their investors are probably looking for some kind of fundraising event,” said Ed Zitron, author of Where’s Your Ed At and host of the Better Offline podcast.
Anthropic said last week it raised $65 billion in a financing round that values the company at $965 billion. The company’s focus is on enterprise customers and developers, with major growth in 2026 fueled by the Claude Code programming tool, which is likely to put it ahead of OpenAI in total value.
The IPO could be a major test of Anthropic’s business model and CEO Dario Amodei, seen here during the World Economic Forum in Davos, Switzerland, in January.
But there are limits to what a company can raise on the private market, and those investors will want to get their money back with a big return. Going public and owning stocks bought not only by individual traders on Robinhood but also by institutional investors such as insurance companies and pension funds can raise huge amounts of money. These companies will also be able to use that stock to get loans that they might not be able to get now.
“The scale of it is impressive,” said Rob Lalka, a business professor at Tulane University. “The fact that this company raised as much money as it did, and spent less money as quickly as it got here.”
Will there be a crash?
The confidential filing means that we do not yet have access to the documents required for the public stock sale, including a prospectus detailing the business’s operations and the risks and challenges it faces.
That public filing will eventually have to include details of the company’s finances and the range of risks it may face. Transparency will also be key. “If they lie to investors, the company can be held liable,” Corrigan said.
Just as companies like Google, Apple, Meta and Microsoft have quarterly calls, where CEOs take questions from investment analysts about the direction of their businesses, Anthropic and its peers will also have to regularly report financial information. The CEOs of Anthropic and OpenAI — Dario Amodei and Sam Altman, respectively — will be subject to the same question.
More importantly, public stock exchanges in major AI-specific companies will put the values of those companies in the hands of investors, including the general public, who can buy and sell based on perceptions of the companies’ movements or the AI industry as a whole.
Watch this: Watch SpaceX Starship Flight Test 12 End With A Bang
If, as some observers suggest, the industry is overstocked, such fluctuations could deflate the bubble — or even inflate it.
Zitron believes it could lead to WeWork-like disruptions, as “people realize how bad the underlying economy is.” WeWork, which operates co-working spaces, filed for an IPO in 2019, with filings that raised serious investor concerns about its corporate governance and ability to make a profit. The company released its filing more than a month later.
It’s hard to know what will happen a year or two down the road, but the future of these companies is often binary, leading to collapse or a major merger. Zitron believes that if OpenAI or Anthropic survives, it could be absorbed or swallowed by a hyperscaler like Google, Microsoft or AWS — or eventually delisted from Nasdaq to serve as a smaller niche service provider.
Wall Street may also decide to ignore negative profit and loss numbers. Lalka pointed to Meta, which spent billions of dollars on the “metaverse” and changed its name from Facebook to reflect the shift to technology it has since abandoned. AI companies can get the same shrug from investors.
“It probably won’t lead to the kind of tough accountability that some say will happen here,” Lalka said.
Some companies, like Uber, have been unprofitable for years. Lalka said that if that remains the case with AI companies, investors may want to focus more on revenue-generating projects than “side requests,” as OpenAI has called functions such as its Sora video production program.
Another possibility is that investors are not chasing the actual value of the company but what they think others will pay for the shares. That possibility, which Corrigan calls a “beauty contest,” may give companies a temporary impression of higher valuations.
“There is a big risk for investors that prices will increase sharply and then come back to Earth,” he said.