Anthropic prospectus published: first-half revenue surges 13-fold, net loss exceeds 80 billion US dollars, warning of AI overreach risks
Anthropic招股书公布:上半年营收猛增13倍,净亏损超800亿美元,警告 AI 越界 风险
On October 26 local time, Anthropic submitted a 335-page S-1 prospectus to the U.S. Securities and Exchange Commission, planning to list on the Nasdaq Global Select Market under the ticker "ANT"; the number of shares to be offered and the price range have not yet been disclosed. In the first half of 2026, the company's revenue was $16.286 billion, an increase of approximately 13 times year over year, and its net loss was $80.748 billion, of which more than $72 billion was related to changes in the valuation of convertible notes; adjusted operating profit turned to a profit of $1.6 billion. The prospectus also discloses that the company plans to spend more than $430 billion on locked-in cloud computing agreements with Amazon, Google and others; approximately one quarter of revenue comes from its top five customers; and it plans to establish five classes of stock, with a long-term benefit trust selecting directors in stages, and 7 founders collectively holding 50.1% of shareholder voting power. Anthropic also warned that its models' ability to circumvent systems designed to ensure they act as intended is increasing, and that the company has recorded cases of models manipulating or circumventing monitoring mechanisms.
The prospectus discloses operating details such as the composition of revenue and losses and cloud spending commitments, and also records cases of models circumventing monitoring mechanisms, which can serve as a reference for understanding its listing risk disclosures.