Anthropic files for potentially the biggest IPO ever as CEO warns of AI's existential risk to humanity
SAN FRANCISCO (AP) — Anthropic, a major artificial intelligence company, filed paperwork on Monday for what could be the biggest stock market launch ever. At the same time, its chief executive is warning that advanced AI could threaten the future of humanity.
The filing prepares the San Francisco company to begin selling shares on Wall Street next month. Its IPO, or first sale of shares to the public, could be bigger than the one SpaceX held in June, which raised a record $86 billion. If Anthropic goes public first, it would beat rival AI company OpenAI to the stock market. Both companies are expected to have very large IPOs.
READ MORE: Anthropic and OpenAI warn about AI safety and seek to influence how it is controlled
The filing comes weeks after CEO Dario Amodei wrote an essay arguing that work on the most advanced AI models should slow down. An IPO could give Anthropic a huge amount of money to expand faster. It is not yet clear how much money the company will raise, how much it will be worth or what its shares will cost. Even lower estimates would make the five-year-old startup one of the world's most valuable companies. The deal could also make Amodei and his sister and co-founder, Daniela, billions of dollars richer.
The 335-page filing shows that Anthropic's revenue is growing quickly, but the company has not yet become profitable. Like many technology companies that go public, it is still losing money. Anthropic, which makes the Claude chatbot, reported a net loss of $80.7 billion in the first half of 2026. However, its losses were much smaller after certain items were left out of the calculation. During the same period a year earlier, the company lost $13 billion.
Private investors valued Anthropic at $965 billion earlier this year. Reports suggest its value could rise to $1.5 trillion or even $2 trillion, although its final value has not been decided. If it is worth $1.5 trillion, Anthropic would rank among the world's most valuable publicly traded companies, alongside Tesla.
Reports say the IPO could raise as much as $100 billion. That would beat the current record, set earlier this year by Elon Musk's company SpaceX.
Anthropic said in its filing with regulators that its shares will trade on the Nasdaq Global Select Market under the symbol "ANT."
Many risks, but a huge financial reward for the founders
Anthropic said in the filing that it could use some of the money to buy other companies and technology. It could also invest in the equipment and systems it needs to expand and compete with its rivals.
Former OpenAI leaders founded Anthropic in 2021. Since then, the company has been competing with OpenAI for control of the AI market. Amodei and OpenAI CEO Sam Altman are widely seen as bitter rivals.
Both companies face many other competitors, including Google, newer startups in the United States and China, and SpaceX, which also has an AI business. Anthropic, OpenAI and SpaceX's AI division have all been spending more money than they earn. Building AI systems requires enormous resources, raising concerns that the industry could be in a financial bubble, where prices rise beyond what companies are really worth.
The need for money may explain why Anthropic is moving quickly to sell shares to the public this year instead of waiting until next year, as OpenAI plans to do. Another concern is that demand for AI stocks could fall sharply or even collapse in the coming months. Investors are becoming more worried about how much money AI companies need. Several technology companies have recently dropped their plans to go public.
The IPO would help Anthropic cover any shortage of cash and could make its seven founders extremely wealthy. Together, the founders own nearly 500 million shares. However, the filing does not make clear what percentage of the company those shares will represent after the IPO.
Forbes estimates that Dario Amodei, 43, and his 39-year-old sister Daniela, Anthropic's president, are each worth $15.5 billion. Forbes ranks them as the 176th and 177th richest people in the world.
The filing also includes unusual rules that reflect concerns about AI's dangers. Amodei, his sister and their five fellow co-founders will control 50.1% of shareholder votes through a special type of share. Anthropic has also appointed independent members to a group called the Long-Term Benefit Trust. These members have no financial interest in the company and have the power to appoint board directors. The trust is meant to make sure the company does not put profits ahead of public safety.
Another unusual part of the filing is its long list of risks. Every public company must explain the risks it faces, but Anthropic's list runs for nearly 100 pages and contains some striking warnings.
Some risks are ordinary business problems, such as losing important employees or facing delays caused by government rules. Others are much more serious. The company warns that people with harmful intentions could use AI models to build chemical weapons. Authoritarian leaders could use the technology to control and punish their citizens. Recent incidents involving AI agents that acted in harmful ways after hacking computer systems have also raised concerns. Anthropic warns that its AI models could "pose catastrophic or existential risks to humanity."
Calls to slow down AI development
The IPO comes as reports show that AI can solve problems beyond human ability and may bring major benefits. But AI can also carry out harmful tasks that threaten lives, raising fears that people could lose control of the technology.
In September, Amodei said the AI industry should slow down its rapid development so that safety measures have time to catch up. He warned that without a slowdown, AI could create large numbers of AI agents that could take control of the entire internet within six to 12 months.
Amodei published a plan on his website calling for stronger checks on the AI industry. He noted that Anthropic is already taking some of these steps itself. Other measures, however, would require AI companies to work together and cooperate with governments around the world, including authoritarian governments.
"I believe that if slowing down bought us even an extra year or two before models reach critical levels of capability, and we used that time to advance alignment, we could greatly reduce the risk that something goes seriously wrong," Amodei wrote.
Disputes with the U.S. government
Anthropic has tried to stand apart from other AI companies by presenting itself as a company that puts safety first. It says it wants to develop AI responsibly rather than grow at any cost with few safety rules.
That position is part of the reason for its dispute with the U.S. government. The conflict began in February, when President Donald Trump and Defense Secretary Pete Hegseth accused Anthropic of putting national security at risk and labeled the company a supply chain risk. This label means the government considers a company a possible danger to its supply system.
In March, Anthropic sued the Pentagon over the label. The company called the government's actions an "unlawful campaign of retaliation" because Anthropic had refused to allow its technology to be used by the military without restrictions. The lawsuit made an already unusual public dispute over the use of AI in war and mass surveillance even more intense.
OpenAI, the company behind ChatGPT, agreed to work with the Pentagon just hours after the government took action against Anthropic over its position. Anthropic also filed a separate, more limited lawsuit challenging a different rule the Pentagon is using in its effort to label the company a supply chain risk.
A judge in California ruled in Anthropic's favor in the first case, while a judge in Washington, D.C., sided with the government in the second. Both sides are appealing the decisions.
Condon reported from New York.