Anthropic Files for IPO, Discloses $80.7 Billion First-Half Loss
Anthropic has reported a loss of $80.7 billion in the first half of the year, almost all of it caused by accounting losses on debt that can be turned into company shares. The company made this disclosure as it prepared to sell shares to the public, after filing its IPO prospectus with the U.S. Securities and Exchange Commission (SEC) on Monday. At the same time, its revenue is growing very quickly. Revenue reached $16.3 billion in the first six months, up from $1.1 billion in the same period a year earlier. Revenue jumped 1,368% in the quarter that ended on June 30 alone. However, the company warns that this rapid growth may slow in the coming months.
During the first six months of the year, Anthropic spent $9 billion on the computer systems and other technology needed to run its AI. That was more than it spent during all of 2025. The company expects this to remain its biggest expense for some time. It has already agreed to spend hundreds of billions of dollars on computing power in the future. These agreements include cloud computing deals with Amazon and Google. Both companies invested in Anthropic early on and provide the infrastructure that runs its Claude AI models.
A valuation between $1.5 trillion and $2 trillion
The share sale is expected to happen within weeks. Anthropic hopes its IPO will become a key measure that Wall Street uses to decide how much major AI companies are worth, including OpenAI. OpenAI filed confidentially for its own IPO in June and is expected to go public by early 2027. Reuters reports that some investors now value Anthropic at around $1.5 trillion, down from $2 trillion a week earlier. The change came after the company shared financial figures at a recent investor event. “No company wants to linger too long on the edge of going public, it’s a distraction. Maybe they’d have gotten more next year, but they decided to rip the band-aid off now,” said Matt Kennedy of Renaissance Capital.
The timing may be difficult. The number of companies planning IPOs this fall has fallen as financial markets have become more unstable. Investors are also growing concerned about the AI investment boom, which is being funded heavily by debt. They are questioning when the trillions of dollars being spent on data centers and computer chips will start to pay off. Still, major technology companies continue to seek large amounts of money. Meta plans to raise up to $95 billion, while SpaceX has just made a $40 billion deal with Apollo to lease computer chips and servers.
More than half of the companies on the Fortune 500 list are already Anthropic customers. Each of these customers generated at least $100,000 in revenue for Anthropic over the past year. However, the company's IPO filing also warns about serious risks. In controlled tests, its most advanced AI models have shown behaviors such as sabotaging computer code, helping with fraud, and even actions “resembling blackmail.” The filing says: “Our increasing use of AI systems to accelerate our own research may cause the capabilities of our models to advance more rapidly than our ability to make them safe.” Anthropic has already faced questions about its safety practices. A former researcher accused Anthropic and OpenAI of “racing straight to self-improving superintelligence.”
The deal’s details
Anthropic was founded in 2021 by former OpenAI executives Dario and Daniela Amodei. The company was valued at $965 billion in May. It plans to list its shares on the Nasdaq stock exchange under the symbol “ANT.” Goldman Sachs and Morgan Stanley will lead a group of 20 banks handling the IPO. The offering follows SpaceX's record-breaking IPO, which raised $75 billion in June and valued the company at $1.75 trillion. Anthropic's prospectus does not yet say how many shares it plans to sell or what price it expects to charge for them.