At a San Francisco investor gathering, Anthropic disclosed that its revenue for the three months ending September rose 45% to $16.8bn, up from $4.7bn in the first quarter and $11.6bn in the second. The company framed the growth as evidence that AI firms can generate substantial earnings despite spending billions on computing infrastructure, a point it emphasized to potential shareholders.
During the same event, Anthropic announced that it had recorded a profit for the third quarter on an adjusted basis, a metric that strips out one-off expenses. The adjusted profit figure was presented as a sign that the firm is on a clear path toward sustainable earnings, and attendees reportedly left the session with a positive outlook on the company’s financial trajectory.
Safety was a recurring theme, with chief executive Dario Amodei urging a slowdown in the pace of AI development. A researcher at the firm warned that there is “greater than a 10% chance that AI could kill all humans” within the next decade, while a former employee said staff were “genuinely frightened” about the speed of progress and its possible consequences for humanity.
Anthropic also reported that it had detected and blocked attempts to misuse its model for malicious purposes, including activities that could facilitate the creation of biological weapons. The firm said these interventions demonstrate its commitment to responsible deployment of powerful AI systems, even as it expands its commercial offerings.
The broader market reacted cautiously to the news, with several US technology stocks slipping after concerns about AI spending. In Asia, Softbank shares fell more than 9%, while South Korean chipmakers Samsung and SK Hynix dropped 9.5% and 11% respectively. Broadcom chief Hock Tan warned that many investors have “no clue” and set “extreme expectations,” a sentiment echoed after inflation data suggested higher borrowing costs could tighten funding for AI projects.
Rival OpenAI has similarly highlighted safety issues, with chief executive Sam Altman stating in September that the company will not pursue a public listing this year while it addresses those concerns. Analysts note that OpenAI’s potential valuation exceeds $1tn, underscoring the high stakes as AI leaders balance growth ambitions with calls for greater oversight.