Writing Team
.png)
OpenAI's chief financial officer Sarah Friar has privately suggested pushing the company's IPO from 2026 to 2027, according to the Wall Street Journal. Her reasoning: OpenAI isn't ready to meet the reporting standards public markets require, and its spending commitments need to be brought under control before it opens its books to public scrutiny.
This is notable not because IPO delays are unusual, but because of who is saying it and what it implies about what she found when she looked closely.
The Revenue Picture Is Not Reassuring
The Journal reported earlier this week that OpenAI missed recent revenue targets. OpenAI is private and not required to disclose financials, so the details are murky — but the company's spokesperson offered a response that raised more questions than it answered: OpenAI has hit its Q1 revenue goals, the spokesperson said, but those goals are different from the ones its investors know about.
That is a genuinely strange thing to say publicly. The company appears to be claiming it hit targets that its own investors weren't tracking. Whether that's spin, miscommunication, or something more concerning is unclear. What's clear is that a company spending money at world-historic scale — committed to $1.4 trillion on data centers over eight years — needs its revenue story to be coherent and credible before it asks public markets to value it.
It currently isn't.
The Numbers That Should Worry Anyone Paying Attention
Internal documents reported late last year projected OpenAI losing $74 billion in 2028 alone. The same analysis showed Anthropic on course to break even that same year, driven by faster-than-expected enterprise revenue growth. Google's recent earnings call reported $20 billion in AI-derived revenue. OpenAI, the company that arguably created the modern AI moment with ChatGPT, is being outpaced on the revenue side by rivals who moved more deliberately on enterprise monetization.
Friar's concern about data center spending makes sense in this context. OpenAI has been committing capital at a pace that requires revenue acceleration to justify. If the revenue acceleration isn't materializing on the timeline the spending assumed, the financial model has a problem that an IPO will make very, very public.
The Friar Factor
The Journal's profile of Friar draws comparisons to Sheryl Sandberg at Facebook and Gwynne Shotwell at SpaceX — executives brought in to install financial discipline around visionary but financially chaotic founders. The comparison is flattering and probably accurate about the dynamic. It's also worth noting the Journal's own footnotes on her track record: Square's stock dropped 10% shortly after she resigned. Nextdoor's stock "tumbled" after the SPAC she oversaw. She also presided over Nextdoor's 2023 layoffs cutting 25% of staff before leaving for OpenAI the following year.
None of that is disqualifying. Complicated outcomes are the norm in tech finance. But it adds texture to the picture of a CFO who has seen what happens when a company goes public before it's ready — and who appears to be trying to avoid repeating that experience at considerably higher stakes.
The IPO Race Nobody Should Be Running
Banks have reportedly told both OpenAI and Anthropic directly that whoever reaches public markets first will get to define the new industry. That framing — first mover captures the narrative — is how investment banks talk when they want to generate urgency around fees. It is not obviously good advice for either company.
Going public before your financial story is legible forces you to manage quarterly earnings expectations while simultaneously trying to build transformative technology on multi-year timelines. The companies that have done this poorly — and there are many of them — spent years distracted by investor relations at precisely the moment they needed operational focus.
Friar is apparently trying to prevent that outcome. Altman, by all accounts, wants to accelerate. That internal tension — between the person responsible for the balance sheet and the person responsible for the vision — is the most honest signal available right now about where OpenAI actually stands.
A company that genuinely had its financial house in order wouldn't have this disagreement in public.
What It Means for the Industry
OpenAI setting the terms of the AI era commercially is no longer the foregone conclusion it looked like eighteen months ago. Anthropic is growing enterprise revenue faster. Google has a functioning AI revenue line. Meta is spending more on infrastructure and reporting profitable quarters. OpenAI is spending more than anyone, missing targets it hasn't disclosed, and debating internally whether it can survive the scrutiny of public markets.
The technology may still be the best. The business, at this moment, is a different question.
For marketers and growth leaders building on OpenAI products and betting on its roadmap, the financial instability of the underlying company is worth factoring into your dependency calculus. Diversifying AI partnerships isn't just a technical decision anymore. Our team at Winsome Marketing helps organizations build AI strategy that doesn't bet everything on one vendor's survival. Let's talk.