Anthropic's IPO Filing Shows $16.3bn Revenue in Half a Year
The S-1 filed on Monday gives the first audited figures for the AI company, including a first-half loss of $80bn driven largely by a one-off charge.
Anthropic has filed to go public on the US stock market. The investor prospectus, an S-1 released on Monday, is the first audited look at the finances of the maker of Claude. It gives no target valuation, although some investors have valued the five-year-old company at between $1.5tn and $2tn.
Revenue and losses
In the first half of 2026 Anthropic brought in $16.3bn and lost $80bn. A $72.4bn one-off charge accounts for much of the loss. For all of 2025 the company reported revenue of $4.585bn and a loss of roughly $33.9bn.
Computing power makes up most of its costs. Spending on AI infrastructure reached nearly $9bn in the first six months of the year, up from $2.3bn a year earlier. Its five largest customers account for 25% of its business, and the filing says the focus is largely on other businesses rather than consumers.
Opportunity and risk
The filing describes the chance of AI reaching the level of "a country of geniuses in a datacenter", and lists biology and medicine, economic growth, global spread of that growth and government services as areas of opportunity. It says the company works with several pharmaceutical and biotechnology firms and has cut the time to write clinical trial reports "from months to minutes".
On risks, Anthropic names four groups: misuse by bad actors, loss of control, authoritarian abuse and economic disruption. It states that its most capable models are capable of providing "meaningful assistance to malicious actors" seeking to develop chemical, biological or other dangerous weapons. Its safeguards, it says, "are incomplete" and may be circumvented or fail.
The prospectus also describes what the company believes was the first documented large-scale cyber-attack with little human intervention. A Chinese state-sponsored group used a Claude coding tool to attempt to infiltrate roughly 30 targets worldwide, including technology firms, financial institutions, chemical manufacturers and government agencies.
Among business risks, Anthropic lists dependence on its top executives, competition from Google and OpenAI, regulation of AI and data centres, and its status as a public benefit corporation. It also points to data centre moratoriums, US immigration policy, and the possibility that it cannot obtain insurance for certain AI-related liabilities.
Source: The Guardian