The news
OpenAI told investors its annualized revenue is approaching $50 billion, the Financial Times reported on October 8, 2026. That is about $20 billion less than a figure that circulated only days earlier, and shares of companies tied to AI spending fell the same day.
According to TechCrunch, an Axios report on September 29 said OpenAI's revenue run rate was nearing $70 billion. TechCrunch, citing the FT, said that higher number came from attempts by OpenAI's own investors to build a direct comparison with Anthropic's annualized revenue. SiliconANGLE put the earlier figure at $68 billion.
The gap appears to come down to accounting. A person familiar with the matter told SiliconANGLE that the earlier figure included gross revenue from partners, to allow a more direct comparison with Anthropic, while the new one is net revenue. TechCrunch noted that Anthropic counts sales made through cloud partners in its figures and OpenAI does not, which makes the two companies hard to compare directly.
OpenAI also pointed to growth. According to SiliconANGLE, the company cited 77% growth in its annual revenue run rate during the third quarter and 107% growth in its enterprise business run rate. TechCrunch said it had contacted OpenAI for comment.
Markets reacted on October 8. SiliconANGLE reported that the Nasdaq fell 1.25%, its worst day since mid-August, and the S&P 500 fell 0.5%. Oracle (ORCL) fell 5.5%, Intel (INTC) 5.3%, CoreWeave (CRWV) 8%, Advanced Micro Devices (AMD) and Broadcom (AVGO) 4% each, Super Micro Computer (SMCI) about 5% and Nvidia (NVDA) 2.9%. The outlet tied the declines to the report; other factors may also have played a part.
The numbers
- OpenAI annualized revenue (per FT)
- approaching $50 billion
- Earlier reported figure
- nearly $70 billion (Axios, per TechCrunch)
- Q3 run-rate growth cited by OpenAI
- 77%
- Enterprise run-rate growth cited by OpenAI
- 107%
- Nasdaq on October 8
- -1.25%
- CoreWeave on October 8
- -8%
Why CEOs should care
For CFOs and procurement teams, the episode is a lesson in definitions. A run rate built on gross partner sales and one built on net revenue can differ by tens of billions of dollars. When a vendor or a vendor's investors quote growth figures, ask whether the number is gross or net, whether it includes cloud-marketplace resale and what period it annualizes.
Boards and investors should look at exposure, not headlines. Many enterprise AI plans assume that leading model makers will keep raising money easily and keep cutting prices. A lower revenue base does not change OpenAI's products, but it can affect how much capital it raises, how fast it builds and what it charges. If a large share of your AI roadmap depends on one vendor, ask what happens to your pricing and capacity if that vendor's funding gets tighter.
CIOs and buyers of AI infrastructure should note the companies that moved most: cloud and chip suppliers such as Oracle and CoreWeave. Their order books depend partly on AI labs' spending. Ask your providers how much of their capacity is committed to a few model makers, and what contract protections you have if their plans change.
The bigger picture
OpenAI and Anthropic are both preparing for public listings, and comparisons between them have become a scorecard for the whole AI sector. TechCrunch reported that OpenAI's IPO has been pushed to early 2027, and Anthropic said on August 17, 2026, according to CNBC, that its annualized revenue reached $65 billion in July. Because the two count revenue differently, the reported figures are not like-for-like.
Holger Mueller of Constellation Research told SiliconANGLE that AI companies need massive funding and it is still unclear how much will be required, which makes their real value hard to judge, and that investors' calculations leave no room for error. The October 8 sell-off suggests that even a correction in how a number is measured can move a large set of stocks tied to AI spending.
What’s next
Watch for OpenAI to give more formal financial disclosure as it moves toward a listing, and for analysts to restate comparisons with Anthropic on a common basis. For buyers, the practical step is to ask vendors to state revenue and usage claims in consistent, audited terms.
What “Fact-checked” means
Fact-checking means testing a story’s facts against the evidence before it is published. This story went through at least two separate checks before this version was published.
- What we checked
- Its names, figures, dates, job titles, quotes and who said what were checked against the story’s sources, including its main source where it could be opened. The headline was checked for accuracy and overstatement.
- How
- A first check reviewed the whole story. If it passed, a second, skeptical check went back to the sources to look for mistakes in the most important facts. If a check flagged the story, it was edited to fix the problems found, and a separate re-check then reviewed the whole story again.
- Who
- The checks are made by our newsroom, as steps kept separate from the writing, under rules set by our editor, Hussein Mukhtar. A story the checks still flag is held for the editor, who decides whether it is fixed, published or dropped.
- If something is wrong
- “Fact-checked” does not mean error-free. If a material error is found after publication, we correct the story and add a note saying what changed. Report an error
