What investors were told
Late on September 13, Reuters relayed a report that Anthropic expected positive adjusted operating profit for a second consecutive quarter. The original article drew on several people familiar with information sent to shareholders. [1 · Reuters]
According to the same report, the company’s gross margin exceeds 80% before accounting for the share of revenue passed to distribution partners and the cost of training models. These exclusions are substantial: the reported margin describes only part of the business’s economics. [1 · Reuters]
What remains unconfirmed
Reuters explicitly stated that it could not independently verify the information. Anthropic did not provide the agency with a comment outside normal business hours, and the report included no public financial statements against which to check the adjustments or cash flows. [1 · Reuters]
It is therefore accurate to describe this as a reported expectation and an adjusted operating metric, rather than proven, sustainable net profit. Subsequent reporting periods should show whether the result holds up after the expenses excluded from the cited margin. [1 · Reuters]
Sources
- Reuters — report on Anthropic’s second profitable quarter — September 13, 2026, 22:54 UTC; the news agency has not verified the information
- Anthropic — company information — Official source; checked September 14, 2026
- Larrain et al. — the effects of companies going public — NBER Working Paper 29219, 2021; international empirical study
- Anthropic — Transparency Hub — Official source; updated July 23, 2026
- Anthropic — Responsible Scaling Policy — Official source; updated August 14, 2026
- IPO study authors — company selection and causal estimation — Authors’ account of the study, September 22, 2021. Checked September 14, 2026; it does not examine Anthropic or explain the company’s performance before a listing.
Expert commentary
If the information is accurate, it is a more meaningful signal than another private-company valuation: a repeated positive operating result suggests that current revenue may cover a substantial share of current expenses. But there is no established confirmation, and the word “adjusted” leaves room for exclusions. I therefore view the report as a sign of improving economics, rather than proof of a completed transition to sustainable profitability. [1 · Reuters]
The improvement may come from the scale of usage: once trained, a model serves many queries, while inference optimization lowers the cost of each additional request. However, training new generations requires large, separate investments, and some revenue goes to distribution channels. The reported margin of more than 80% is calculated before these items, so it cannot be compared directly with the gross margin of a conventional software company. [1 · Reuters]
For competition, what matters is not the percentage itself but the ability to convert demand into funding for the next cycle of research and infrastructure. Sustainably positive operating economics would reduce Anthropic’s dependence on new funding rounds and strengthen its position ahead of a possible listing. An alternative explanation is a temporary combination of large contracts, deferred expenses or favorable adjustment methods; without disclosure of the composition of revenue and costs, it is impossible to distinguish between these scenarios. [1 · Reuters] [3 · Larrain et al.]
For customers, greater financial resilience could reduce the risk of an abrupt service shutdown and expand the resources available for support. The other side is that efforts to protect margins could lead to changes in pricing, limits and the allocation of computing resources across service tiers. The quality of customer relationships will show in predictable prices, transparent limits, accessible support and whether model performance offsets the full cost of evaluation and deployment. [1 · Reuters] [2 · Anthropic]
It is important to keep the research basis narrow here. In the study by Larrain and coauthors, a simple comparison of completed and withdrawn IPOs showed no profitability advantage; a positive estimate emerged after accounting for selection using an instrumental variable. This does not prove that a stock market listing automatically makes a business profitable. Anthropic’s current figures relate to the period before a possible listing, so they cannot be explained by an IPO effect. Easing financial constraints remains only a mechanism in a conditional future scenario. [3 · Larrain et al.] [6 · IPO study authors]
The story should be tested against several observable measures: official financial statements, the definition of adjusted profit, training expenses, the share paid to partners, cash flow and whether the result persists for at least several periods. Spending on safety matters separately: Anthropic says safeguards should strengthen as model capabilities grow. A favorable scenario requires that higher profitability is not achieved by cutting those checks or worsening terms for users. [1 · Reuters] [4 · Anthropic] [5 · Anthropic]