They left the slide up after the meeting. You could see the numbers in the corner—revenue short of the mark, more red than anyone expected. I remember thinking: whoever steadies that ship just changed the game.
I will keep this tight. You already know the players—Sam Altman, Sarah Friar, Microsoft, Anthropic, the Wall Street Journal—and you want to know what a delayed IPO would mean for them and for you.
On a conference-room whiteboard: missed targets — Why OpenAI’s CFO is quietly pushing for 2027
You’ve read the Journal pieces by Berber Jin and Corrie Driebusch: Sarah Friar has been scrutinizing OpenAI’s outflows and reportedly told colleagues to consider waiting until 2027 for an IPO. The claim is simple on its face: the company isn’t ready for the reporting rigors of public markets.
Friar’s résumé gives that counsel weight. She ran finance through Square’s IPO and navigated Nextdoor’s SPAC listing—both moves that left scars and lessons. Investors and reporters have compared her to the kinds of executives who tame overambitious founders: Sheryl Sandberg at Facebook and Gwynne Shotwell at SpaceX.
When could OpenAI IPO?
Reports say 2026 was the plan; now the ask is to push to 2027. That delay isn’t theatrical dithering. It’s a plea for cleaner books, steadier revenue, and a slower burn of capital before answering to public shareholders.
At the data-center command post: massive commitments versus revenue reality — How big the gap is
A spreadsheet on someone’s laptop showed a cliff: enormous capital commitments that outpace current revenue streams.
Internal documents reported late last year placed OpenAI’s data-center commitments at $1.4 trillion (€1.29 trillion) over eight years and forecast a $74 billion (€68 billion) loss in 2028 alone. By comparison, Google’s AI-related revenue was reported recently at $20 billion (€18 billion)—a figure that makes potential investors think twice about who’s paying for what.
Anthropic, OpenAI’s biggest rival, has been building enterprise sales fast enough that it could reach break-even sooner. Banks have told both companies bluntly: whoever reaches the market first shapes the industry. That pressure has fed a spending spree that, to me, looks unsustainable without clearer revenue engines.
OpenAI is a pressure cooker, and that pressure is what Friar wants to release before opening the lid.
Why would a CFO delay an IPO?
Because going public changes the rules overnight. Public companies face quarterly scrutiny, stricter accounting, and shareholder expectations that punish mismatch between promises and receipts. A CFO can see the runway and the storm clouds; delaying can preserve valuation and reputation.
In the investor bullpen: internal goals, external pressure — What the tug-of-war means for Microsoft and markets
On calls with banks and partners, the language has been blunt: winner-take-all, get to market first, define the industry. That’s how bankers framed the race to IPO, according to reporting.
OpenAI has pushed back publicly. A spokesman told the Journal that the company hit its first-quarter revenue goals and that it tracks internal targets investors may not see. Microsoft, a major backer and cloud partner via Azure, is watching too—its balance sheet and go-to-market muscle matter to OpenAI’s ability to monetize tools like ChatGPT.
Friar, by many accounts, has started reining in data-center expansion and signaled caution to Altman’s acceleration. She has become the brakes on a speeding car, and that friction could rearrange strategy: slower public timing, firmer enterprise sales, and tighter cost controls.
If you follow corporate drama, this is where names matter: Sam Altman’s optimism, Sarah Friar’s conservatism, Microsoft’s stewardship, Anthropic’s enterprise push, and banks’ blunt timelines.
I’ll say this plainly: an IPO delayed by a year can be salvation or squandered momentum depending on execution. You’ll want to watch three things closely—enterprise revenue growth, capital spending on data centers, and any tighter disclosure on internal targets to investors.
Which do you think will matter most to the market when OpenAI finally files: cleaner math or faster growth?