Anthropic, the San Francisco-based artificial intelligence startup behind the Claude chatbot, has officially filed for a US stock market listing. This highly anticipated public offering, trading under the symbol ANT, is expected to rank among the largest debuts in history and serves as a critical test for the financial sustainability of the current multibillion-dollar AI boom. A leader in the race to create powerful artificial intelligence models and a bellwether of the American AI boom, Anthropic offers the chatbot Claude as its flagship product, which has text generation features as well as advanced coding and cybersecurity capabilities. Which have been the object of intense speculation in recent months as questions of a potential AI bubble swirl, anthropic’s audited filings will offer the first public glimpse of its financial health.
Founded in 2021 by CEO Dario Amodei following his departure from OpenAI, the company has attracted significant investor interest, with market valuations reaching over $1.5 trillion. The company’s investor prospectus emphasizes its core philosophy, stating, “We believe AI is one of the most consequential technologies in human history.” To support its expansion, Anthropic had raised $141.1 billion by the end of June 2026.
The financial disclosures reveal a period of explosive growth coupled with substantial costs. Anthropic reported revenue of $41 million in 2023, which surged to $386 million in 2024. By 2025, revenue hit $4.585 billion, marking an increase of more than 1,000%. In the first six months of 2026 alone, the company generated $16.286 billion. However, these gains are offset by heavy expenditures, with the company disclosing a net loss of $80 billion for the first half of 2026 and total debt of $6.2 billion, which includes convertible notes issued to Amazon Web Services. However, investing in Anthropic and the development of a model that the company claims could revolutionize knowledge work across industries comes with a great deal of risk.
Customer adoption remains a pillar of the company’s strategy. By late September, Anthropic served more than 14,000 enterprise clients. This includes 11,900 customers generating at least $100,000, 1,900 generating $1 million or more, and approximately 250 enterprise accounts exceeding $10 million in revenue. Executives noted that “the speed and depth of our enterprise customer adoption are powerful evidence of Claude’s position at the frontier of AI.” Anthropic researchers and executives have repeatedly warned of AI’s potential to cause human extinction.
Despite this momentum, the company identified significant business risks in its filing, noting that a “substantial portion” of its revenue relies on a concentrated group of top-tier clients. While no single customer accounts for more than 10% of the business, the top five customers provide a quarter of total revenue. Furthermore, the company faces external challenges, including a ongoing dispute with the Trump administration regarding the use of its models in warfare, which led to the firm being designated a supply chain risk, effectively barring it from military contracts. Anthropic listed more than 90 pages of potential risks to its business. This prevents any military agency from contracting with Anthropic.
Security concerns are another prominent feature of the firm’s disclosures. Executives categorized risks into four areas: malicious use, loss of control, authoritarian interference, and economic disruption. Anthropic acknowledged that its most advanced models possess capabilities that could assist bad actors in developing chemical, biological, or other dangerous weaponry. While the firm utilizes technical safeguards, it conceded that these measures are “incomplete” and may be circumvented. The company deploys technical safeguards, “but such safeguards are incomplete, may be outdated in light of evolving threat models, have been and may in the future again be circumvented, or may fail to operate as intended.”.
The company also disclosed details regarding a major security incident in September 2025. Staff documented what was described as the first large-scale, automated cyber-attack without significant human intervention. Anthropic detected a state-sponsored hacking group from China that manipulated the Claude Code tool to attempt infiltrations into roughly 30 global targets, including financial institutions, major tech firms, and chemical manufacturers.
Looking toward the future, the prospectus uses ambitious language to describe the potential of AI agents—semi-autonomous systems that perform tasks for users. The firm characterizes this evolution as “something without historical precedent: what we refer to as a country of geniuses in a datacenter, available to every organization and sector of the economy.” The filing contains some of the hyperbolic language characteristic of AI companies predicting the miraculous potential of their technology.
