Anthropic's "Profitability" Swindle
We recorded 18 claims from this post. 4 were checked against evidence. 14 can’t be checked. 6 more are held. Read the post at its source.
Claims from this post (18)
In the order they appear in the post. Each quotation is the speaker’s words in this post. A label such as “Supported” or “Can’t be checked” is the claim’s current assessment; its page has the evidence and the reasons.
Assessment: Can’t be checkedSaid
“That’s right —-- Anthropic is possibly going to be EBITDA profitable for a single quarter, on a non-GAAP basis.”
What we tested: Anthropic is possibly going to be EBITDA profitable for a single quarter, the second quarter of 2026, on a non-GAAP basis.
Assessment: Can’t be checkedSaid
“While I wouldn’t say this is cooking the books, it’s definitely a shiatsu-grade massaging of the numbers.”
What we tested: Zitron would not call Anthropic's profitability figure cooking the books, but says it is definitely a heavy ('shiatsu-grade') massaging of the numbers.
Assessment: Can’t be checkedSaid
“Anthropic has deliberately leaked a quarterly “profit” where it knows it can suppress its costs, specifically made sure that the journalist gave it the out of “costs might increase,” and released it on the day of NVIDIA’s earnings as a means of keeping the AI bubble inflated.”
What we tested: Anthropic deliberately leaked a quarterly 'profit' for a quarter where it knew it could suppress its costs, made sure the journalist gave it the out that 'costs might increase', and released it on the day of NVIDIA's earnings as a means of keeping the AI bubble inflated.
Assessment: Can’t be checkedSaid
“If Anthropic paid full-rate for its compute in those two months, its economics would shift back to what they’ve always been per my reporting from last year on its AWS costs — a business that has costs that linearly increase with its revenue growth.”
What we tested: Without the discounted compute rate in May and June 2026, Anthropic would again be a business whose costs rise linearly with revenue.
Assessment: Can’t be checkedSaid
“I also severely doubt that Anthropic managed to make the cost of running its services profitable in the space of six months.”
What we tested: It is very unlikely Anthropic made running its services profitable in the past six months.
Assessment: Can’t be checkedSaid
“For large enterprises, Anthropic is taking prepayment of tokens —-- say, $50 million intended to be spread over 12 months that it takes in as revenue.This would both inflate revenue numbers and depress costs, because Anthropic wouldn’t have actually provided the compute necessary to earn that revenue yet.”
What we tested: Zitron guesses Anthropic may be booking enterprise token prepayments as revenue up front, which would inflate revenue and depress costs.
Assessment: Can’t be checkedSaid
“It may very well be booking this up-front.”
What we tested: Anthropic may be booking discounted prepaid Claude credits as revenue up front.
Assessment: Can’t be checkedSaid
“Anthropic could be front-loading annual commitments of any kind —– subscriptions to Claude, enterprise or team agreements, and so on.”
What we tested: Anthropic could be front-loading annual commitments as revenue.
Assessment: Can’t be checkedSaid
“Anthropic could have ratcheted down training to ease the burden on its infrastructure to provide inference.”
What we tested: One possible explanation is that Anthropic could have ratcheted down training to ease the burden on its infrastructure to provide inference.
Assessment: Can’t be checkedSaid
“Nevertheless, the revenue side is where the real problems lie.”
What we tested: The real problems with Anthropic's profitability claim are in its revenue figures.
Assessment: ContradictedSaid
“At this point it’s impossible to reconcile the 2025 reporting with that $5 billion number.”
What we tested: Anthropic's reported 2025 revenue can't be reconciled with its CFO's sworn statement of revenues exceeding $5 billion to date.
What happened: The declaration gives a cumulative lower bound without an upper bound. A larger cumulative sum incorporating the reported 2025 revenue can therefore satisfy the statement; the asserted numerical impossibility does not follow. (U.S. District Court, Northern District of California / CourtListener, )
Assessment: UnresolvedSaid
“While I acknowledge that Anthropic has grown significantly, that level of stratospheric growth does stretch the limits of credibility. Moreover, the fact that previous ARR figures are inconsistent with the leaked charts from Anthropic further raises questions about the credibility of any numbers from the company.”
What we tested: Anthropic's implied growth is hard to believe, and its earlier ARR figures don't match its leaked charts, which casts doubt on all of its numbers.
Assessment: UnresolvedSaid
“In July, Anthropic will start paying SpaceX $1.25 billion a month, - or $15 billion a year, - on top of all of its other compute deals with Google, Amazon and Microsoft.”
What we tested: Starting in July 2026, Anthropic will pay SpaceX $1.25 billion a month ($15 billion a year) for compute, on top of its other compute deals.
Assessment: Can’t be checkedSaid
“There’s also a very compelling argument that Anthropic’s costs will increase and will eat up that profitability, to once again quote the Wall Street Journal:”
What we tested: Zitron says there is a very compelling argument that Anthropic's costs will increase and will eat up its reported profitability.
Assessment: Can’t be checkedSaid
“I will give Dario Amodei credit: nobody does financial engineering and a press-led information war better than Anthropic.”
What we tested: Anthropic is unmatched at financial engineering and waging information wars through the press.
Assessment: Can’t be checkedSaid
“Let me speak directly and with more empathy than usual: if you want Anthropic to win, you should be just as skeptical of these numbers as I am.”
What we tested: People who want Anthropic to win should be just as skeptical of its reported numbers as Zitron is.
Assessment: Can’t be checkedSaid
“Anthropic timed this leak to focus on a specific quarter where it artificially suppressed costs, and gave you the flimsiest proof imaginable, specifically-crafted for you to share it as a triumph and spread the idea that “AI labs are actually profitable,” when their core economics haven’t changed.”
What we tested: Anthropic timed this leak to focus on a specific quarter where it artificially suppressed costs, giving the flimsiest proof imaginable, crafted for people to spread the idea that AI labs are actually profitable, when their core economics haven't changed.
Assessment: UnresolvedSaid
“Costs increase linearly with revenue, and will continue to do so in perpetuity.”
What we tested: AI labs' costs rise linearly with revenue and will keep doing so indefinitely.
Held from this post (6)
These records did not pass every check, so nothing about them is published. Held is not a finding.
- The two reviews read the claim differently
- The criteria are ambiguous. Our reading, not assessed: Anthropic's costs appear to scale linearly with its revenues, except, suspiciously, in Q2 2026.
- The two reviews read the claim differently
- The two reviews read the claim differently
- The record does not meet the current rules. Our reading, not assessed: A sweetheart SpaceX deal, framed as a ramp-up, let Anthropic artificially lower its costs.
- The record does not meet the current rules. Our reading, not assessed: Anthropic's reported profit doesn't show that AI or inference is profitable or that a viable business model is emerging.