OpenAI Is Totally Cooked

What a s**tshow.

Photo by Codioful (Formerly Gradienta) on Unsplash

By now, we all know that OpenAI is up a waterway of excrement without any means of propulsion. From their models being about as useful as a chocolate tea cosy that gives you psychosis to their ever-growing crushing financial losses, things are not looking good for Altman’s plagiarism machine company. But in recent weeks, the thumbscrews have begun to tighten on this saga, and the natural conclusion of this company is finally coming into focus. You see, the colossal flow of cash OpenAI desperately needs to stop itself from imploding is starting to evaporate, and soon this vanguard of the AI bubble could be consumed by its own cash burn. That’s right, OpenAI is in a far worse position than you think.

Microsoft Pullback

Let’s start with the quite frankly hilarious revelation that Microslop is moving to ditch the monster they created.

As reported by the Financial Times and Windows Central, Microsoft’s AI lead, Mustafa Suleyman, has suggested they are shifting away from their reliance on OpenAI. He said, “We have to develop our own foundation models, which are at the absolute frontier, with gigawatt-scale compute and some of the very best AI training teams in the world.” This implies they have realised specialist multi-model AI tools might be better than the general slop machines Altman has pushed.

To say this is surprising is a huge understatement. Microsoft was pivotal in the growth of OpenAI as an early investor, pumping tens of billions of dollars into the startup, securing lucrative deals with OpenAI and acquiring a 27% stake in their for-profit arm. This is why almost all of Microsoft’s AI products, like Copilot and its AI coding tools, are entirely based on OpenAI models. It is also why Microsoft is shoving AI down users’ throats so hard that many are fleeing to Mac or Linux — because the company needs people to use the AI they bet so heavily on. It is absolutely in Microsoft’s interest to support and collaborate closely with OpenAI rather than compete with them. Changing stance like this implies Microsoft acknowledges that it made a mistake on an unfathomable scale here.

This is utterly horrific news for OpenAI in three ways.

Firstly, there is the reputational damage this will cause. Part of what made OpenAI the vanguard for the AI bubble and the largest AI company out there was the financial support, platforming and associated legitimacy given to them by Microsoft. Microsoft turning their back on OpenAI could pull the rug on this concept and severely damage OpenAI’s perceived value. And OpenAI needs to keep its value skyrocketing to ensure investors are happy to hand over the fistfuls of cash the company requires to stave off bankruptcy.

The second is the constant investments. Microsoft was one of the biggest participants in OpenAI’s corporate bailout in 2024 (along with Nvidia and Softbank) and injected billions into it to prevent it from going bankrupt (read more here). OpenAI’s losses are growing, and to prevent bankruptcy, it will need larger corporate bailout-style investments in the future. If Microsoft turns its back on OpenAI, securing these kinds of funds might become a lot more difficult.

But the third issue is the joint financial ties. You see, OpenAI has to share 20% of its revenue with Microsoft, which sounds like a substantial drain. But, in return, Microsoft gives OpenAI heavily discounted cloud computing (The Information reports as much as 75% being discounted) and shares 20% of the revenue it generates selling repackaged OpenAI models, such as Copilot. We don’t know how much Microsoft is paying OpenAI for their end of the revenue share bargain, but thanks to Zitron, we know that “OpenAI’s inference spend with Microsoft Azure between CY2024 and Q3 CY2025 was $12.43 billion.” This means that the financial deals between OpenAI and Microsoft are likely heavily skewed to OpenAI’s benefit and could be saving them tens of billions of dollars a year overall.

So, if the relationship between Microsoft and OpenAI breaks down, it could dramatically increase OpenAI’s costs, making their already dire financial situation even worse.

And it isn’t just Microsoft stabbing Altman in the back.

NVIDIA Pullback

Et tu, Huang?

I have already talked about this topic, but it bears repeating. Back in September 2025, Nvidia announced a plan to invest $100 billion into OpenAI in the form of computing power. While this was rightfully perceived as a giant red flag for circular funding, it was nonetheless a critical deal for OpenAI, as it desperately needed funds (or equivalent resources) at this scale. According to the Wall Street Journal, this deal has since stalled and effectively fallen through. Apparently, OpenAI wanted to seal the deal within a few weeks, but it turned out Nvidia wasn’t happy with their business model. Other reports have cited different concerns, and it appears Nvidia made a counteroffer of a $20 billion investment instead, but this also appears to have gone nowhere.

This was supposed to be a well-needed slam dunk. Nvidia is very close with OpenAI, and this was a circular deal that would greatly benefit them. The fact that such an important deal has failed to materialise is, quite frankly, damning. If OpenAI can’t land these easy funds, how on Earth will it raise the colossal sums required just to keep going?

We will discuss exactly how much they need soon because it isn’t just OpenAI struggling to raise funds but those associated with OpenAI, too.

Oracle Blowback

Oracle is one of the main middlemen in the AI bubble. They don’t develop AI models like OpenAI or build the hardware on which they run, like Nvidia. Instead, they are a cloud computing company, so their responsibility is to buy the hardware and effectively lease it to AI companies.

Cloud computing hardware isn’t cheap, which means Oracle has always had a significant amount of debt on its books to pay for its enormous infrastructure. But AI requires far larger, more capable, and more expensive hardware, so when Oracle began to shift to supplying the AI industry, its debt exploded. Over the course of 2025, its total debt increased by more than 33% to around $100 billion! This monumental increase in debt spooked investors, who saw it as a warning sign of a bubble, causing bond insurance prices for Oracle’s debt to skyrocket (read more here).

Still, Oracle has gone “in for a penny, in for a pound” and is doubling down. They recently announced that they plan to raise $50 billion in debt in 2026 to pay for AI data centres. This announcement increased Oracle’s stock by 2%, as investors saw it as a sign of strong AI infrastructure demand. After all, Oracle wouldn’t just raise such a huge amount of debt to build infrastructure without strong assurances that the client would use it. Right?

Well, then it happened. Oracle tweeted, “The Nvidia-OpenAI deal has zero impact on our financial relationship with OpenAI. We remain highly confident in OpenAI’s ability to raise funds and meet its commitments.” This comment shattered the narrative. It strongly implies that OpenAI’s demand is the primary driver for this $50 billion debt. But it also highlights the main problem: OpenAI is struggling to raise the investment it needs, as we can see with Nvidia. So, rather than comforting investors, it spooked them, causing a rapid drop in stock price, with one VC likening the frenzy to a “bank run”.

This is quite a clear demonstration that the market is not confident in OpenAI’s ability to raise cash — which, again, shouldn’t be especially difficult, as Microsoft and Nvidia were some of the biggest investors in OpenAI over the past few years, and both are seemingly distancing themselves from OpenAI now.

The True Scale Of The Mess

Okay, so how much does OpenAI need to raise to prevent collapse? It’s all well and good to point out that OpenAI’s backers are turning heel and running, but just how far is OpenAI down the excrement waterway?

Well, OpenAI has committed to $1.4 trillion worth of compute expenses by 2033!

HSBC analysts found that OpenAI will have to fulfil $792 billion of these commitments by 2030. However, it also found that it would still be far from profitability by 2030, and even if it secured all the potential investment deals it was offered (such as that $100 billion Nvidia deal), it would still be $207 billion short. This implies that OpenAI still needs to find another $608 billion to pay for its expenditure commitments from 2030 to 2033.

Keep in mind, this is just to meet OpenAI’s spending commitments. It doesn’t include covering OpenAI’s seismic annual losses, which are growing rapidly. I estimated they could be over $15.6 billion for 2025, which is triple their 2024 net loss.

So, in reality, that $100 billion deal from Nvidia was a drop in the ocean compared to what OpenAI requires. They need an additional $700 billion plus just to pay for their committed expenditure between now and 2033. If losses keep growing at their current rate, OpenAI will need to raise substantially more on top of that just to pay off their losses and avoid bankruptcy. They depended on the Nvidia deal to be the first domino to fall, setting the stage for other massive investments.

Not to mention that a lot of these expenses aren’t some distant concern for Altman. Ed Zitron found that OpenAI needs to raise some $400 billion this year to complete its promised data centre deals. Admittedly, Zitron and many others have pointed out how impossible, nonsensical and hyperbolic these deals are, but this still highlights a major problem. OpenAI doesn’t have anywhere near the money it needs! And, if OpenAI fails to fulfil this dramatic expansion, it will crush companies like Oracle, which has bet everything on OpenAI doing so. When this happens, it will only highlight the financial folly of the AI bubble, and everything will come crashing down.

It really shouldn’t be a surprise that even mainstream journalists like The New York Times’ Sebastian Mallaby are beginning to predict that OpenAI will go bankrupt in the very near future.

Summary

The walls are closing in on OpenAI. Its major backers are getting cold feet, all while it needs exponentially more cash to be shoved into its black hole to keep the financial charade of the AI bubble from collapsing and to stop itself from imploding due to its rapidly growing losses. This can only end one way, and it is insane that people are trying to convince you otherwise.

---

Thanks for reading! Everything expressed in this article is my opinion, and should not be taken as financial advice or accusations. Don’t forget to check out my YouTube channel for more from me, or Subscribe. Oh, and don’t forget to hit the share button below to get the word out!