“AI startups are an artificial source of revenue.”
Our reading:
AI startups are an artificial source of revenue for AI labs because they pay with venture capital rather than cashflow.
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
The claim's testable core is that AI startups, as a group, pay OpenAI and Anthropic mainly with venture funding rather than with operating cash flow. The target period is 1 January to 8 September 2026. The packet has no aggregate or representative data on how AI startups funded their model spending in that period. The only evidence item is a PYMNTS report available on 4 September 2026 (e1). It says Cursor and GitHub Copilot drove about $1.2 billion of Anthropic's $5 billion revenue milestone in the prior year, and that GitHub is owned by Microsoft. That figure is for 2025, not the target period. It also does not describe any buyer's cash flow or funding sources. The evidence neither supports nor contradicts the claim.
Caveats
- 'Artificial' is a characterization. Only the claim that the payments depend on venture funding can be tested.
- The packet has no cash-flow, funding, or margin data for any AI startup in 2026.
- The only evidence item covers 2025 and names just two large customers of Anthropic. It says nothing about OpenAI's startup customers.
- The packet does not show what share of the labs' revenue comes from AI startups as opposed to other customers.
Evidence (1 item)
Context only: PYMNTS,
Coding tools Cursor and GitHub Copilot alone drove roughly $1.2 billion of the company’s $5 billion revenue milestone last year, close to a quarter of total revenue from just two customers
Why it matters: Shows that big token buyers include both a VC-backed startup (Cursor) and a Microsoft product (Copilot). It refers to 2025 and says nothing about startups' cash flow, so it is context only.
- Where
- paragraph 5
- Publisher
- PYMNTS
- Kind of source
- News report
- 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
- 'Artificial' is a characterization. The checkable core is that AI startups' payments to OpenAI and Anthropic are funded mainly by venture capital rather than by operating cash flow.
- Read as a general claim about AI startups as a group, as the speaker frames it ('in virtually every case I've found'), not as a claim about every single startup.
- He describes an ongoing state with no named period. The target period is 2026 up to the post date, the period the post's examples (Harvey's 2025 to 2026 raises) point to.
Right if: Evidence that most AI startups paying for model access run at operating losses or negative cash flow, and fund their token spending from venture raises.
Wrong if: Evidence that a substantial share of AI-startup spending with the labs comes from startups that generate positive operating cash flow, or that pay without depending on new venture funding.
We wrote this test after the outcome was known.
What would settle it
Representative data, filings, or credible reporting for January to September 2026 that set AI startups' operating cash flow and venture funding against their spending on OpenAI and Anthropic models.
We found 37 claims in this post. See them all.
Full record
Source and dates
Where's Your Ed At, Concentration Risk, paragraph 37.
- 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 2026-09-08.
Where it was said (1)
Said once in the posts we read.
: Where's Your Ed At, Concentration Risk
AI startups are an artificial source of revenue.
- Where
- paragraph 37
- Published
- 2026-09-08
- 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
A reviewer said our test wasn’t fair to what was actually said.