Apparently they overstated revenue in an attempt to try to provide a direct comparison with Anthropic's reported metrics.
From the article: "According to a person with knowledge of the matter, the discrepancy arose from attempts by OpenAI’s own investors to produce a direct comparison with Anthropic’s annualised revenues. The pair calculate the figure in different ways, with Anthropic including the revenue from sales via cloud partners such as AWS and Google Cloud, while OpenAI does not.
Efforts to “gross up” OpenAI’s annualised revenue led to reports that the group’s annualised revenue had hit $40bn in August. The company has since told investors its revenues have grown more than 70 per cent, leading to the $70bn figure"
**sorry the gift link can only be viewed 3 times..
>The pair calculate the figure in different ways, with Anthropic including the revenue from sales via cloud partners such as AWS and Google Cloud, while OpenAI does not.
gift link didn't work for me, and is this poorly phrased? because it seems implausible that OpenAI doesn't typically include revenue from their models being used on AWS. Perhaps the "gross up" is referring to how the number is included? like Anthropic was using the value pre-removal of revenue sharing and putting the revenue share subtraction as a separate expense?
[not a finance guy so someone tell me I'm wrong if that's not a plausible reading]
Obviously not the same thing as lying, but Anthropic had also been juicing the revenues with making smart models incredibly verbose. In August my org 3x more in API credits vs July. In September the spend returned to July levels partially because they made models less verbose, but mostly because we've changed how we are using them.
All that tells you is that they haven't at all abandoned the disinformation: Now they are blaming Anthropic for it and trying to shift attention to them.
You can still learn something from it: Look at what they do, not what they say - look at how sophisticated their public communication is. They deliver that information in the perfect manner - not only the redirection and striking a blow against their rival, but they use an anonymous "person with knowledge of the matter": A named source at OpenAI might betray the self-interest in the statement, but some anonymous third party is just reporting what they know.
These guys are very good at it, though that shouldn't surprise you. Look at their product, in one sense a highly effective disinformation machine.
This is an interesting shift compared to the past where OpenAI would’ve been public a long time ago (due to various regulations) so we would have much more direct insight.
Right now we have a ~$1 trillion company which a ton of the “economy” and valuations are based on, with near zero information on how it’s doing.
I think the fact that we have so little information is the most important information we have. If OpenAI had a sound business plan and was on strong financial footing, they'd have IPO'd.
All of this continued stalling and obfuscation can only mean one thing, IMO: OpenAI has no long-term viability and they're desperately hoping for some new breakthrough to reinvent their business model before the VC money faucets turn off for good.
> OpenAI has no long-term viability and they're desperately hoping for some new breakthrough to reinvent their business model before the VC money faucets turn off for good.
If a startup is riding a hype cycle and is one of two leaders in the global industry with unreal growth numbers, they can IPO whenever they want. The incentives lean toward doing an IPO before the hype runs out, not delaying it.
If they were worried about running out of VC money, going to the public (the P in IPO) would be the move.
Companies don't actually have to go public quickly or even at all, even though that's been drilled into us as the only goal of every investor-backed startup.
> If they were worried about running out of VC money, going to the public (the P in IPO) would be the move.
Yes? They were geared up for IPO this year until pushing it back. See all the marketing shenanigans around solving mathematics for this month's flavour. They have a balancing act to manage between the hype and the reality of the business.
They do because the other part of the equation is that they need to keep spending a lot of money to build out infrastructure faster than their two most significant competitors, one of which is public and wildly profitable (for non-AI reasons) and another has already filed for IPO. So it puts them at a disadvantage to walk away from a massive cash infusion.
I don't know why comments pointing out this simple fact are getting downvoted.
The oversimplified view that has been drilled into startup discussions for years has been that IPO is the singular goal for every startup and they need to get there as fast as possible, but that hasn't been true for a long while. There are high profile examples like Stripe with no intent to go public any time soon. Some public companies are even gradually doing share buybacks partially to remove their public exposure.
Being a public company kind of sucks in many ways. I'll admit my sample size is small, but every post-IPO CEO I've known has expressed some regrets about going public. It was a fascinating revelation to me after being raised on the idea that IPO is the ultimate victory goal of every startup.
> I think the fact that we have so little information is the most important information we have. If OpenAI had a sound business plan and was on strong financial footing, they'd have IPO'd.
Anthropic likely would not be saying, in October, that they planned to go public next month, if this were also true of their business.
In the last ~month, OpenAI announced a delay to its IPO and Anthropic put a relatively near-term range on its IPO date. These are very different signals.
Those numbers intentionally exclude the single largest operating expense that Anthropic has: model training. [1]
So yeah, if they stop training models forever, Anthropic will probably start making a profit... until someone else with better models comes along to eat their lunch.
Or if model training is more of a rollercoaster, where spending gets you to the top of the hill where you create a massive internal model which can then build the next version of itself for cheaper and cheaper amounts relative to human R&D costs. If Anthropic is first over that hill, they can race far ahead.
They have found a path to “profitability” iif you define “profitable” in a way that makes every early stage start-up that has at least one paying customer as “profitable”. Literally any start-up has a COGS lower than their income, but that doesn't mean anything at actual profitability given that the rest of their expenses dwarfs it.
I believe the entire basis of their profitable quarter was getting a discount on compute from Musk.
All these figures are so utterly weaselly. AAR is a made up measure to make them look good. If they cannot show GAAP numbers, they are hiding something. Full stop. While as private companies they are under no legal obligation to show us their books, their PR and intent to go public requires it.
I think we're going to suddenly see them greatly scale back training and try to sell inference-only, but they all know when they do that someone can jump up and outstrip them.
But only as long as training actually improves models significantly. As soon as those improvements stay below a certain threshold, the better move is to invest your R&D money into other things like harnesses or new tricks one can play with existing models and the immense cost of training is just not worth it to be 0.5% ahead.
I'm absolutely certain that we will reach that point, just not when. Could come sooner than we think though.
The reason is that companies can choose the best timing to go public - when their financial look the best - and they do.
Anthropic trying to go public very soon is a good tell their financial look pretty decent. OpenAI postponing the IPO is a very good tell theirs look bad.
There are bits and pieces of info scattered everywhere but no coherent picture. We know from a16z report [1] that only 2% of US households pay for AI subscriptions, so most of the seat based pricing comes from business and enterprise agreements. We know that OpenAI ads business has $1 billion USD in annualised revenue run rate [2]. We know from OpenRouter data [3] that in 2025 70% of the API token spend was across all the proprietary models (Anthropic did very well in 2025, while OpenAI clawed back market share in 2026).
It all starts to look like a very low margins business, and reminds me very much of telecom industry.
What regulation makes it so that a large private company would have to go public? You have it backwards. Increased regulations make it more burdensome for a company to go public, driving companies like OpenAI to remain private. If you made it less expensive and burdensome to IPO (decreased regulation), you would see companies go IPO earlier.
This also coincides with a growing market for private credit and VC which certainly helps companies stay private for longer.
There was a rule[1] pre-2012 that forced public disclosure (akin to listed companies) for private companies when it had >500 shareholder (which counted employees with shares). This made it so that companies had a choice to stay private with all obligation of public disclosure or go public for added benefit of tapping public market.
In 2012 this was relaxed in JOBS Act which relaxed the 500 threshold to 2000 but more importantly it ignored employees so now private companies of gargantuan trillion dollar valuation and thousands of employees have no disclosure requirements.
So, this is a classic case of regulation that did well but was relaxed and now creates hidden risks.
I think they would have kept it under 500 if they had to. I doubt this is the determining factor. In fact a lot more than 2k investors have exposure through SPVs or holding companies on top of holding companies. So no, I don't think this was the determining factor that allowed OpenAI to stay private longer.
Regulations that limit the ability of investors to invest in private companies, although these have been weakened in recent decades, which helps fuel the growth of private credit markets that allow private companies to stay private.
I'd say it's the growth of private markets to allow companies to keep getting funding even at the $100 billion range while staying private that has fueled the trend to stay private rather than SOX and other new regulations for public corporate governance dissuading them from going public.
>Regulations that limit the ability of investors to invest in private companies, although these have been weakened in recent decades, which helps fuel the growth of private credit markets that allow private companies to stay private.
it really is a privatize the gains socialize the losses situation, isn't it? due to the new rules (or lack thereof), public investors didn't have access to all that growth.
no way it ever gives you a return like, say, the amazon IPO could've.
Losses are much more privatized staying private. Instead of hitting people's 401k or pension fund, this is mostly contained to a concentrated set of VC and PE investors, not large public markets.
> This is an interesting shift compared to the past where OpenAI would’ve been public a long time ago (due to various regulations) so we would have much more direct insight.
There's a good recent YouTube video about the shift in regulations that switched IPOs from being a way to raise money for growth to being a way to dump on retail investors after all the significant growth has been funded by private investors:
How much money do they have left? It is hard for me to see how OpenAI doesn't fail at this point. There is no business, no moat. Honestly, the best outcome seems like failing up into a Microsoft acquisition at this point.
Any company that would hypothetically acquire them would need to be able to fully - and indefinitely - subsidize their unsustainable operational costs. I don't see how that's realistic even for a company like Microsoft.
it will fail the day there is a downturn of the economy. That day will be at a 2001 or a 2008 like event - anytime within the next few years (I've no crystal ball, but strong convictions haha).
And yes, they will be acquired by a company which will have survived the next crash at a fraction of their currently estimated valuation and we will truly have the next ride of the economy .. many years ahead if 2001 is an example.
Why does it have to be such a crash? Could it be just flatlining for a long time? Or perhaps slowly going down? What makes you so sure that it will be a big booom like dotcom or 2008?
There is a moat: government contracts. Everything from NSF grant reviews, drone warfare, DHS visa processing and Medicare/Medicaid claims processing are up for grabs under the right administration.
The entire US economy is propped up by this bubble, so they will be propped up as much and as long as possible. Basically all money these days is going into this charade, to stop or even slow it down would cause a disastrous collapse
The headline should've been "OpenAI annualised revenues $20B less than previously signalled by us". The FT is just reporting high number to create a story, then a low number to create another story.
> far short of the $70bn reported by the FT and other media outlets late last month based on information that was provided to investors.
Media was mislead by second hand information and misled the public, now they are 'shocked' they reported incorrectly..
Still, seems it is still true that their number is not directly comparable to Anthropic's because they calculate it differently, I think that part still stands and is pretty relevant here.
Sorry Bence I trust the FT journalists to have due diligence w.r.t. what evidence of AR they got from investors (who got it from OpenAI). I completely believe that the company helmed by "not consistently candid" sama bullshits investors about their AR which is the point of this article.
Though I agree with your sentiment that FT is reporting this stuff in a way to stir the pot and create outrage. Speculating about a private company like this is stupid.
> based on information that was provided to investors.
It was OpenAI spreading their bullshit annualized revenue.
OpenAI and Anthropic always play this silly game to pretend they are in anyway viable. It is always ARR, "adjusted" revenue, etc. "We are profitable when we pretend we don't have expenses".
How does that contradict what parent said? Yes, obviously they are pointing at something for their numbers, but something as vague as "based on information that was provided to investors" might still just be entirely nonsense and is certainly not enough to establish confidence as to the validity of the claim.
Parent implied that the previous number was a fabrication from FT, instead of it being a fabrication from OpenAI.
The press that reports on this shit is very much complicit, they report on bullshit metrics spread by these companies to generate hype.
> outside of OpenAI control.
If OpenAI is as uncertain of their numbers to the magnitude of 20B, they should stop spreading bullshit metrics. In fact this should be considered fraud.
What I was saying: What FT presents in the way they do should not be enough to move the needle and convince anyone that this happened, regardless of what might have happened and that it could have happened.
People can just write stuff. That does not make it wrong but that also does not make it right. If your claim rests on some claim that some anonymous actor got some information, that's just not super convincing and neither is pointing at it as some sort of truth.
>> based on information that was provided to investors.
>It was OpenAI spreading their bullshit annualized revenue.
Did you read the article?
>The discrepancy arose from attempts by OpenAI’s own investors to produce a direct comparison with Anthropic’s annualised revenues, according to a person familiar with the matter. The pair calculate the figure in different ways, with Anthropic including the revenue from sales via cloud partners such as Amazon’s AWS and Google Cloud, while OpenAI does not.
>Investors’ efforts to “gross up” OpenAI’s annualised revenue prompted reports that the figure was around $40bn in July, said the person.
>OpenAI later told its backers that its annualised revenues had jumped more than 70 per cent since July, prompting reports that the figure was about $70bn at the end of September — a number the company did not deny.
>However, the new investor presentation shows close to $30bn annualised revenues in July.
Sounds like what happened wasn't that openai "spreading their bullshit annualized revenue", it was that they gave some vague figure that investors the media and other investors extrapolated, and it turned out that extrapolation was incorrect. Both the $40bn and $70bn figure did not come from openai directly.
I'm waiting for the IPO; I was hoping we'd see less news like this prior. I'm not sure if plain shorting, or puts are the correct action; I suspect the former, as timing the latter is not reliable.
judging from SpaceX performance, I think the market can handle a couple more low float high valuation issuances
The public markets have floated multiple names up to $6tr marketcap/valuation
so the debuts at $1tr valuation from the private markets don’t mean anything
pre-IPO investors will just collar like they did SpaceX, nullifying any price volatility for them at the high share price, and giving them all the liquidity collateral they need for borrowing
This is pretty much the main point of his frustration.
The world has standardised methods of accounting. Not only do Anthropic and OpenAI avoid using those methods, they both use the same phrase “annualised revenues” to describe two radically different accounting processes.
They’re both also leaking those annualised numbers slowly to the press at irregular intervals, which hints that they’re disclosing new numbers in the days after a big sale lands. So you see “$30bn annualised” because they managed to land a $1bn contract the week before, bumping the annualised figure up by $12bn compared to the start of the previous month, and the end of the next.
Zitron complained about a lot of things, one some he was wrong (eg. llm are not useful), on others(eg. "magic accounting" or datacenters ) he has very good points ... but we will see.
Apparently they overstated revenue in an attempt to try to provide a direct comparison with Anthropic's reported metrics.
From the article: "According to a person with knowledge of the matter, the discrepancy arose from attempts by OpenAI’s own investors to produce a direct comparison with Anthropic’s annualised revenues. The pair calculate the figure in different ways, with Anthropic including the revenue from sales via cloud partners such as AWS and Google Cloud, while OpenAI does not. Efforts to “gross up” OpenAI’s annualised revenue led to reports that the group’s annualised revenue had hit $40bn in August. The company has since told investors its revenues have grown more than 70 per cent, leading to the $70bn figure"
**sorry the gift link can only be viewed 3 times..
gift link didn't work for me, and is this poorly phrased? because it seems implausible that OpenAI doesn't typically include revenue from their models being used on AWS. Perhaps the "gross up" is referring to how the number is included? like Anthropic was using the value pre-removal of revenue sharing and putting the revenue share subtraction as a separate expense?
[not a finance guy so someone tell me I'm wrong if that's not a plausible reading]
Just to clarify from my understanding of the quote, "they" here is openai investors, not openai.
I'm confident both companies are lying about their revenues.
You can still learn something from it: Look at what they do, not what they say - look at how sophisticated their public communication is. They deliver that information in the perfect manner - not only the redirection and striking a blow against their rival, but they use an anonymous "person with knowledge of the matter": A named source at OpenAI might betray the self-interest in the statement, but some anonymous third party is just reporting what they know.
These guys are very good at it, though that shouldn't surprise you. Look at their product, in one sense a highly effective disinformation machine.
Right now we have a ~$1 trillion company which a ton of the “economy” and valuations are based on, with near zero information on how it’s doing.
I think the fact that we have so little information is the most important information we have. If OpenAI had a sound business plan and was on strong financial footing, they'd have IPO'd.
All of this continued stalling and obfuscation can only mean one thing, IMO: OpenAI has no long-term viability and they're desperately hoping for some new breakthrough to reinvent their business model before the VC money faucets turn off for good.
If a startup is riding a hype cycle and is one of two leaders in the global industry with unreal growth numbers, they can IPO whenever they want. The incentives lean toward doing an IPO before the hype runs out, not delaying it.
If they were worried about running out of VC money, going to the public (the P in IPO) would be the move.
Companies don't actually have to go public quickly or even at all, even though that's been drilled into us as the only goal of every investor-backed startup.
I'm not certain OpenAI or Anthropic have a viable business, either, but Spacex definitely pulled a massive scam.
Yes? They were geared up for IPO this year until pushing it back. See all the marketing shenanigans around solving mathematics for this month's flavour. They have a balancing act to manage between the hype and the reality of the business.
OTOH, if a company has a sound business plan and strong financial footing it may not need to IPO -- unless the founders or VCers want out ASAP.
The oversimplified view that has been drilled into startup discussions for years has been that IPO is the singular goal for every startup and they need to get there as fast as possible, but that hasn't been true for a long while. There are high profile examples like Stripe with no intent to go public any time soon. Some public companies are even gradually doing share buybacks partially to remove their public exposure.
Being a public company kind of sucks in many ways. I'll admit my sample size is small, but every post-IPO CEO I've known has expressed some regrets about going public. It was a fascinating revelation to me after being raised on the idea that IPO is the ultimate victory goal of every startup.
Whatever private-market liquidity events they will be permitted to participate in will be highly disadvantaged compared to the other two groups.
What about Stripe?
In the last ~month, OpenAI announced a delay to its IPO and Anthropic put a relatively near-term range on its IPO date. These are very different signals.
So yeah, if they stop training models forever, Anthropic will probably start making a profit... until someone else with better models comes along to eat their lunch.
[1] https://www.morningstar.com/news/marketwatch/2026091414/the-...
All these figures are so utterly weaselly. AAR is a made up measure to make them look good. If they cannot show GAAP numbers, they are hiding something. Full stop. While as private companies they are under no legal obligation to show us their books, their PR and intent to go public requires it.
Remember a few weeks ago when all the AI labs said "we need to slow down, to uh, prevent destroying the world"?
I'm absolutely certain that we will reach that point, just not when. Could come sooner than we think though.
The reason is that companies can choose the best timing to go public - when their financial look the best - and they do. Anthropic trying to go public very soon is a good tell their financial look pretty decent. OpenAI postponing the IPO is a very good tell theirs look bad.
It all starts to look like a very low margins business, and reminds me very much of telecom industry.
[1] https://www.a16z.news/p/state-of-markets-ii
[2] https://www.reuters.com/business/media-telecom/openais-ad-bu...
[3] https://openrouter.ai/state-of-ai
This also coincides with a growing market for private credit and VC which certainly helps companies stay private for longer.
In 2012 this was relaxed in JOBS Act which relaxed the 500 threshold to 2000 but more importantly it ignored employees so now private companies of gargantuan trillion dollar valuation and thousands of employees have no disclosure requirements.
So, this is a classic case of regulation that did well but was relaxed and now creates hidden risks.
[1]: https://www.investopedia.com/terms/5/500-shareholder-thresho...
I'd say it's the growth of private markets to allow companies to keep getting funding even at the $100 billion range while staying private that has fueled the trend to stay private rather than SOX and other new regulations for public corporate governance dissuading them from going public.
Which regulations are these?
no way it ever gives you a return like, say, the amazon IPO could've.
Losses are much more privatized staying private. Instead of hitting people's 401k or pension fund, this is mostly contained to a concentrated set of VC and PE investors, not large public markets.
There's a good recent YouTube video about the shift in regulations that switched IPOs from being a way to raise money for growth to being a way to dump on retail investors after all the significant growth has been funded by private investors:
https://www.youtube.com/watch?v=roe3SgezmmU
https://www.reuters.com/legal/transactional/openai-targets-3...
And yes, they will be acquired by a company which will have survived the next crash at a fraction of their currently estimated valuation and we will truly have the next ride of the economy .. many years ahead if 2001 is an example.
Still, seems it is still true that their number is not directly comparable to Anthropic's because they calculate it differently, I think that part still stands and is pretty relevant here.
Though I agree with your sentiment that FT is reporting this stuff in a way to stir the pot and create outrage. Speculating about a private company like this is stupid.
> based on information that was provided to investors.
It was OpenAI spreading their bullshit annualized revenue.
OpenAI and Anthropic always play this silly game to pretend they are in anyway viable. It is always ARR, "adjusted" revenue, etc. "We are profitable when we pretend we don't have expenses".
Its comically bad how this circus is playing out.
The press that reports on this shit is very much complicit, they report on bullshit metrics spread by these companies to generate hype.
> outside of OpenAI control.
If OpenAI is as uncertain of their numbers to the magnitude of 20B, they should stop spreading bullshit metrics. In fact this should be considered fraud.
People can just write stuff. That does not make it wrong but that also does not make it right. If your claim rests on some claim that some anonymous actor got some information, that's just not super convincing and neither is pointing at it as some sort of truth.
>> based on information that was provided to investors.
>It was OpenAI spreading their bullshit annualized revenue.
Did you read the article?
>The discrepancy arose from attempts by OpenAI’s own investors to produce a direct comparison with Anthropic’s annualised revenues, according to a person familiar with the matter. The pair calculate the figure in different ways, with Anthropic including the revenue from sales via cloud partners such as Amazon’s AWS and Google Cloud, while OpenAI does not.
>Investors’ efforts to “gross up” OpenAI’s annualised revenue prompted reports that the figure was around $40bn in July, said the person.
>OpenAI later told its backers that its annualised revenues had jumped more than 70 per cent since July, prompting reports that the figure was about $70bn at the end of September — a number the company did not deny.
>However, the new investor presentation shows close to $30bn annualised revenues in July.
Sounds like what happened wasn't that openai "spreading their bullshit annualized revenue", it was that they gave some vague figure that investors the media and other investors extrapolated, and it turned out that extrapolation was incorrect. Both the $40bn and $70bn figure did not come from openai directly.
The public markets have floated multiple names up to $6tr marketcap/valuation
so the debuts at $1tr valuation from the private markets don’t mean anything
pre-IPO investors will just collar like they did SpaceX, nullifying any price volatility for them at the high share price, and giving them all the liquidity collateral they need for borrowing
The world has standardised methods of accounting. Not only do Anthropic and OpenAI avoid using those methods, they both use the same phrase “annualised revenues” to describe two radically different accounting processes.
They’re both also leaking those annualised numbers slowly to the press at irregular intervals, which hints that they’re disclosing new numbers in the days after a big sale lands. So you see “$30bn annualised” because they managed to land a $1bn contract the week before, bumping the annualised figure up by $12bn compared to the start of the previous month, and the end of the next.
2.5bn in revenue for all of July. That is a disaster.
Squeaky bum time
Anyway, if I had a hundred bucks to burn, I’d bet this is a move to undermine Anthropic’s IPO.
So much ARR.
My anuallized revenue is about 4.5M. I just need now to get a salary every day.
And the sad fact is even though they did this, they were still 20 billion too low.
Not only will this never be an IPO, this is a signal of a collapse of the economy