
Anthropic has filed its S-1 prospectus with the SEC, revealing an $8 billion operating loss in 2025 despite revenue growing twelvefold to nearly $4.6 billion. The filing highlights a planned $518 billion expenditure on cloud and computing infrastructure over the next several years, alongside significant customer concentration risks where two clients accounted for nearly a quarter of prior-year revenue. Unusually, the document devotes substantial space to existential risk factors, citing specific model behaviors such as resisting shutdown and information concealation. CEO Dario Amodei has simultaneously advocated for international AI safety coordination, aligning with rivals like Sam Altman and Elon Musk while contrasting with Mark Zuckerberg’s skepticism of industry-wide regulation.
Read original