A small startup has put credit ratings on two of the world’s best-known AI companies, and the results are getting attention. Forbes reported on October 12, 2026 that AIR Platforms, an AI-powered bond-rating startup, gives Anthropic an investment-grade rating of BBB and rates OpenAI at BB+, which counts as junk. These are unsolicited ratings from a young firm, and the article itself says to take them with “a heap of salt”. Note that Anthropic is the company behind this site’s AI assistant, so we are sticking strictly to what was reported.
What AIR Platforms Said?
According to Forbes, AIR published its unsolicited report on October 1, 2026. The investment-grade cutoff is BBB-, so Anthropic’s BBB sits just above it, while OpenAI’s BB+ sits just below. The major ratings agencies have not weighed in. AIR is a 10-person New York startup that has raised $6.1 million, led by CEO Glenn Carvajal, who spent 12 years at Moody’s. See background on credit ratings on Wikipedia and the original story on Forbes.
The Ratings at a Glance
| Item | Detail |
|---|---|
| Anthropic rating | BBB (investment grade) |
| OpenAI rating | BB+ (junk) |
| Investment-grade cutoff | BBB- |
| Estimated EBITDA margin | Anthropic 5.4%; OpenAI negative 33.3% |
| Tech benchmark median margin | 43.3% (seven large tech firms) |
| Rater | AIR Platforms, unsolicited, not hired by either company |
| Company responses | OpenAI and Anthropic did not respond to requests for comment |
Why the Ratings Differ?
AIR says Anthropic’s positive adjusted operating result was the main reason it rated higher than OpenAI, and that the BBB relies on recent results suggesting losses have narrowed. For comparison, Forbes notes S&P ratings of AAA for Microsoft, AA+ for Alphabet and Apple, and BBB- for Oracle. Ratings matter because they shape borrowing costs: the article cites Apple’s 10-year notes yielding 5.4% against 7.5% for Oracle’s. Follow similar coverage on our OpenAI page.
Why It Matters?
Both companies are raising billions to pay for chips and data centres, and Forbes notes, citing the Financial Times, that Morgan Stanley and Goldman Sachs have been talking with S&P, Moody’s and Fitch about investment-grade ratings for them after their IPOs. A good rating lowers the cost of debt, so these opinions are being watched even though they come from a small, unofficial source.
Big Caveats
AIR has not seen either company’s full financial statements. It uses public information and AI to estimate the missing pieces, and its co-founder reportedly said that incomplete public information is the biggest constraint and small changes in the numbers could shift the picture quickly. Its grades do not directly include off-balance-sheet items such as leases, computing contracts and guarantees, and the article says the assessment may be generous. Forbes also cites large commitments on both sides, including at least $518 billion in future infrastructure commitments for Anthropic and a forecast $856 billion in data-centre and chip spending for OpenAI by decade’s end.
What Is Not Confirmed?
These points are not confirmed:
- Any rating from S&P, Moody’s or Fitch, none of which has been issued.
- The accuracy of AIR’s estimates, since it lacks full financial statements.
- How off-balance-sheet obligations would change the results.
- Whether either company seeks a formal rating.
- The companies’ own views, as neither responded.
Quick Answers
What rating did Anthropic get? BBB, investment grade, from AIR Platforms.
And OpenAI? BB+, which is junk.
Are these official ratings? No. They are unsolicited ratings from a startup.
Have big agencies rated them? Not yet, according to Forbes.
Source: Forbes report (October 12, 2026) and Forbes post on X. Ratings and figures are as reported and are AIR Platforms’ unsolicited estimates.
