“They report that Anthropic hit $19 billion in annualized revenue and OpenAI hit $25 billion in annualized revenue — which has been confirmed to refer to a 4-week-long period of revenue multiplied by 12 — as proof that the AI bubble is real, ignoring the fact that both companies lose billions of dollars and that my own reporting says that OpenAI made billions less and spent billions more in 2025.”
Our reading:
OpenAI and Anthropic both lose billions of dollars, and OpenAI made billions less and spent billions more in 2025 than reported.
Assessment: Unresolved
What happened
Was it true for the period described?
Outcome: Unresolved
We checked, and the evidence so far doesn’t settle it.
Judged on a plain reading. They named no measure, so we used the reading a typical informed reader would take. It is spelled out under “How we judged it”.
Evidence up to . Checked .
Why
Ars Technica reported on June 16, 2026 that OpenAI's 2025 net loss was nearly $39 billion, including a large nonrecurring accounting charge [e2]. This supports the OpenAI loss component [e2]. Epoch AI reported on February 4, 2026 that Anthropic projected a $5.2 billion EBITDA loss for 2025; this does not establish final net losses [e3]. CNBC reported on February 20, 2026 that OpenAI's 2025 revenue was $13.1 billion and cash burn was $8 billion [e1]. The supplied reports do not identify the earlier reports being challenged or establish comparable revenue and spending discrepancies of at least $2 billion each [e1, e2]. The evidence therefore leaves the compound claim unresolved [e1, e2, e3].
Caveats
- The evidence does not identify a definitive earlier baseline for either claimed OpenAI discrepancy [e1, e2].
- Cash burn and accounting expenses are different measures; the packet does not reconcile them [e1, e2].
- Anthropic's projected EBITDA loss does not establish its final 2025 net loss [e3].
- Ars Technica provides secondary reporting on financial statements; the underlying statements and Financial Times verification were not directly inspected for this packet.
Evidence (3 items)
Context only: CNBC,
OpenAI generated $13.1 billion in revenue in 2025
Why it matters: This supplies a possible earlier realized-revenue baseline. The adjacent spending figure is explicitly cash burn, so it cannot be compared directly with total accounting expenses.
- Publisher
- CNBC
- Kind of source
- News report
- Measures
- OpenAI, Annual revenue, USD billions, estimate, 2025-01-01 to 2025-12-31
- About events on
- Published
Context only: Ars Technica,
nearly $39 billion in 2025.
Why it matters: The report supports the OpenAI loss component and reports revenue near the earlier CNBC figure. It describes audited statements reviewed by the Financial Times, but this remains secondary access and does not reconcile the compound claim’s spending baseline.
- Publisher
- Ars Technica
- Kind of source
- News report, audited
- Measures
- OpenAI, Annual net loss, USD billions, company reported, 2025-01-01 to 2025-12-31
- About events on
- Published
Context only: Epoch AI,
spending at each of these labs is 2-3x higher than revenue.
Why it matters: For Anthropic, the article describes a projected $5.2 billion EBITDA loss, based on press reporting. That supports financial strain but is neither verified final net loss nor evidence about OpenAI’s spending discrepancy.
- Publisher
- Epoch AI
- Kind of source
- Secondary summary
- Measures
- Anthropic, Projected annual EBITDA loss, USD billions, estimate, 2025-01-01 to 2025-12-31
- About events on
- Published
How we judged it
- Kind of claim
- Observation
- Deadline
- None given. We don’t add one.
Where we had to interpret
- Assess the supplied proposition about losses and discrepancies, rather than the surrounding annualized-revenue figures.
- Use calendar 2025 for the explicit annual comparison and the latest completed annual period for the ongoing loss assertion.
- Read billions as a difference of at least $2 billion, separately for revenue and spending.
- The comparison requires an identifiable earlier report measuring the same annual revenue and spending.
- The financial figures are reporting-based estimates. The speaker does not specify whether losses mean net losses, operating losses or cash burn; the ordinary net-loss reading must not be populated with another measure.
Right if: Comparable records establish multibillion-dollar annual losses at both companies and show OpenAI’s 2025 revenue at least $2 billion below, and spending at least $2 billion above, the identified earlier reports.
Wrong if: Comparable records establish that either company did not have multibillion-dollar losses, or that either claimed OpenAI discrepancy was absent or below $2 billion.
We wrote this test after the outcome was known.
What would settle it
Independently verified 2025 net-loss figures for both companies, the earlier OpenAI reports being challenged, and comparisons on matching accounting bases showing whether revenue was at least $2 billion lower and spending at least $2 billion higher.
We found 8 claims in this post. See them all.
Full record
Source and dates
Where's Your Ed At, The Beginning Of History, paragraph 79.
- Said
- We read the source
- live page, full text, public
- Evidence up to
- Checked
We read this wording on 2026-10-28. The post may have been edited after 2026-03-10.
Where it was said (1)
Said once in the posts we read.
: Where's Your Ed At, The Beginning Of History
They report that Anthropic hit $19 billion in annualized revenue and OpenAI hit $25 billion in annualized revenue — which has been confirmed to refer to a 4-week-long period of revenue multiplied by 12 — as proof that the AI bubble is real, ignoring the fact that both companies lose billions of dollars and that my own reporting says that OpenAI made billions less and spent billions more in 2025.
- Where
- paragraph 79
- Published
- 2026-03-10
- We read it
- , live page
- Access
- full text, public
History
Versions
- Current: Unresolved. Evidence up to 2026-10-28.
What happened to this record
PublishedBy us
Published after both AI reviews and the final check passed.
Held backBy us
The final check failed: the write-up said more than the evidence showed.
Held backBy us
The final check failed: a cited source didn’t say what the write-up said, and the write-up said more than the evidence showed.