Anthropic Guides to Second Straight Positive Quarter on $11.5 Billion in Revenue
Anthropic told shareholders its adjusted operating income will be positive for a second straight quarter after recording an adjusted operating profit in the second quarter, the Financial Times reported on Sept. 13.
Second-quarter revenues were $11.5 billion, up 14-fold from a year earlier, the Financial Times reported.
On Saturday, Sept. 12, Chief Executive Dario Amodei publicly called for a slowdown in artificial intelligence in a lengthy essay. Politico reported the essay, while Reuters reported that Amodei urged AI companies to slow model development.
Anthropic introduced Claude Opus 5 on July 24. The company described the release as a step-change improvement for the Opus tier powering long-running agents with improvements in coding and professional work. Its subsequent listings include "Our position on open-weights models" on July 27 and "Investigating three real-world incidents in our cybersecurity evaluations" on July 30. Anthropic
In August, Anthropic announced that Mariano-Florentino (Tino) Cuéllar will join as Chief Global Affairs Officer on Aug. 4. Its listings for the month also include "How Claude's text watermark works" on Aug. 14, "Funding better evaluations of AI's impact on wellbeing" on Aug. 25 and "Expanding our support for scientists" on Aug. 27. Anthropic
On Sept. 1, the company listed "Developing Enterprise Frontier Safeguards with our customers." It also said it is opening a research preview of the Model Hardware Standard, a shared specification for AI agents to safely operate physical devices, to scientific research labs and advanced manufacturers. Anthropic
The broader context here is operating leverage arriving alongside a heavier governance load. The math is stark. Fourteen-fold growth to $11.5 billion with two consecutive positive prints on an adjusted operating basis suggests inference efficiency, enterprise pricing, or utilization gains are offsetting compute intensity for now. The qualifier matters. Adjusted operating income excludes items that still affect cash, dilution, and balance-sheet risk, so the quality of the adjustment and the margin structure on long-running agent workloads will determine persistence.
In my view, the sequencing is what finance desks will probe. A shareholder update guiding to sustained adjusted profitability was reported one day after the chief executive's public call to slow model development. That combination puts focus on capital allocation discipline, contracting terms for agent deployment, and how safeguards, watermarking, evaluation funding, and hardware interfaces are priced and provisioned. For readers tracking private AI financials, the question is no longer scale alone. It is unit economics under longer-horizon tasks and the overhead required to deploy them inside enterprise and scientific environments.
