Last Updated on July 7, 2026 by Taya Ziv
On its tax return, OpenAI declared its mission “unconstrained by a need to generate financial return.” That was the whole point. A nonprofit building artificial general intelligence for the benefit of humanity. No shareholders to appease. No quarterly earnings calls. Just the mission.
This week, the company confidentially filed paperwork for what could be the largest IPO in history. Target valuation: north of $852 billion. Investment banks: Goldman Sachs and Morgan Stanley. Target date: September.
The nonprofit that didn’t need money is about to ask the public market for it.
The Numbers Nobody Wants to Say Out Loud
Here’s what OpenAI’s own internal projections tell us about the company about to go public.
Revenue in 2025: roughly $13 billion. Projected losses in 2026: $14 billion. That’s not a typo. The company will lose more money this year than it made last year. Total spending in 2026 clocks in at around $22 billion.
But those are the small numbers.
OpenAI has signed $1.4 trillion in infrastructure commitments over the next eight years. That’s trillion with a T. Almost $100 billion of that is earmarked for backup data center capacity alone. The cumulative cash burn through 2029 is projected at $115 billion.
The company is currently spending $3.30 for every $1.00 it makes.
And the pitch to public investors? “Trust us, we’ll be profitable by 2029 or 2030.”
The Faith-Based IPO
Every major tech company that went public had something OpenAI doesn’t.
Google was printing money from AdWords. Facebook had 1.2 billion monthly users and had just figured out mobile ads. Even Amazon, the poster child for “we’ll be profitable later,” was growing revenue 20% per year with positive operating cash flow in most quarters by its IPO.
OpenAI is going public with a 57% loss ratio that it expects to maintain through 2027. Its path to profitability requires the company to grow from $13 billion to $200 billion in annual revenue by 2030. That’s a 15x increase in four years while simultaneously reducing costs enough to flip the P&L.
This isn’t an IPO in any traditional sense. It’s a belief system priced as equity.
The investors buying OpenAI shares aren’t making a financial calculation. They’re making a theological one: does AGI arrive before the money runs out? If yes, $852 billion will look like a bargain. If no, this is the most expensive nonprofit in human history.
The Musk Lawsuit Was the Starting Gun
The timing isn’t subtle. On May 20, a federal jury dismissed Elon Musk’s lawsuit that accused Sam Altman of stealing a charity. Days later, the IPO machinery started moving.
Altman testified that when Musk walked away from OpenAI in 2018, the company was “left for dead.” Eight years later, it’s worth more than JPMorgan, Visa, and ExxonMobil. And the man who tried to kill it in court just involuntarily cleared the legal path for it to go public.
That’s not strategy. That’s narrative arcs that Hollywood screenwriters would reject for being too on-the-nose.
What This Actually Means for Founders
If OpenAI pulls off this IPO, it rewrites the rules of what “fundable” means.
Think about what just got validated: you can lose $14 billion a year, have no clear path to profitability, commit to $1.4 trillion in future spending, and still be worth more than most countries’ GDP. The only requirement is that your mission is big enough, and your losses are denominated in figures so large that people stop calling them losses and start calling them “investment in the future.”
That’s not a business model. That’s a permission structure.
And if you’re a founder raising a Series A right now, here’s the uncomfortable implication: the bar for what constitutes “ambitious enough” just moved so far that your $5 million ARR SaaS might actually be the thing VCs consider too small to matter. When OpenAI is burning through your entire funding round every 45 minutes, the psychological frame of venture changes.
But the other side is darker. If OpenAI goes public and the stock craters because public market investors aren’t as patient as SoftBank, the correction won’t just hit OpenAI. It’ll hit every AI startup in the ecosystem. Public market skepticism is contagious. We saw it with WeWork. We saw it with Peloton. When the biggest players prove that AI economics don’t work at scale, the smaller players lose their funding narratives overnight.
The $200 Billion Question
OpenAI’s entire bull case rests on a single assumption: that AI models will become so essential to every business process on Earth that the company can charge enough to justify the infrastructure.
The company projects $200 billion in annual revenue by 2030. For context, that’s more revenue than Netflix, Uber, Airbnb, Snap, Spotify, and Pinterest make combined. In four years.
Where does that money come from? Right now, the AI companies actually making money are the ones embedding into developer workflows, not the ones selling chatbot subscriptions. Anthropic just posted its first operating profit. Cursor hit a $50 billion valuation. The winners are the ones who’ve figured out how to bill $200/month instead of $20/month.
OpenAI’s bet is that it can do both: be the infrastructure layer AND the application layer. Be AWS and Salesforce simultaneously. That’s never worked before. But then again, no company has ever had this much money to try.
The Paradox Every Investor Should Name
Here’s the thing nobody in the financial press is saying directly.
If OpenAI achieves AGI, the $852 billion valuation is cheap. Absurdly cheap. A system that can do any cognitive work better than humans would be worth more than the entire S&P 500. Buying it at $852 billion would be like buying Google in 2004.
But if OpenAI doesn’t achieve AGI, it’s a company spending $22 billion a year to offer a chatbot that’s marginally better than six free competitors.
There is no middle ground in this investment thesis. There’s no “it’ll be a nice profitable company either way.” The whole pitch is: either it changes civilization, or it’s the most spectacular write-off in financial history. Wall Street has a long history of calling binary bets “strategy” when the numbers are big enough.
And retail investors are about to be invited into that bet in September.
That’s the real story. Not the IPO. Not the valuation. The question of what happens when you take a nonprofit’s mission, wrap it in a trillion dollars of commitments, and sell shares to people who think they’re buying a software company.
