OpenAI's CFO wants to pump the brakes on a 2026 IPO

The investment bankers are pushing for a 2026 blockbuster listing, but the person who actually watches OpenAI's bank account wants to hit the brakes.

According to a report from The Wall Street Journal, OpenAI Chief Financial Officer Sarah Friar has privately advised delaying the company's initial public offering until 2027. Her reasoning is simple: after taking a hard look at the startup's massive spending commitments, she cautioned that OpenAI simply isn't ready to meet the rigorous accounting and reporting standards demanded of public companies.

Trillion-dollar checks and missing targets

The pushback comes as OpenAI faces mounting financial scrutiny. Earlier reporting revealed the private startup had missed its recent revenue targets.

An OpenAI spokesperson pushed back against that narrative, claiming the company did in fact hit its revenue goals for the first quarter of the year. But the spokesperson added an eyebrow-raising caveat: those internal targets are known only inside the company and differ from what investors were told.

Meanwhile, the cash burn is historic. Internal documents from late last year revealed OpenAI committed to spending an eye-popping $1.4 trillion on data centers over eight years. In 2028 alone, the company is projected to lose $74 billion.

By contrast, rival Anthropic is tracking to break even by 2028, largely fueled by rapid enterprise revenue growth. Big Tech is also turning up the heat; Google recently posted a blowout quarter, showing off $20 billion in AI-related revenue.

The adult in the room pushes back

Wall Street banks have been whispering in the ears of both OpenAI and Anthropic, framing the IPO race as a high-stakes, winner-take-all sprint where whoever lists first gets to define the market.

Yet Friar seems unwilling to let bankers dictate the timeline. The Journal compared her role to Silicon Valley's classic steadying hands, like Sheryl Sandberg at Meta or Gwynne Shotwell at SpaceX.

Friar previously steered Jack Dorsey's payments firm Square through its IPO and later took Nextdoor public, where she also navigated a brutal 25% staff layoff in 2023 before jumping to OpenAI in 2024. While Altman keeps slamming the accelerator on infrastructure, Friar is actively working to rein in cash outflows until real revenue catches up.

Why it matters

We think taking the foot off the IPO accelerator is arguably the smartest and most responsible move OpenAI could make right now.

Investment bankers want their underwriting fees, and Altman wants endless billions for compute. But going public while burning toward a projected $74 billion annual deficit, all while hiding behind secret internal targets that differ from investor expectations, would be an invitation to public market carnage.

For enterprise customers and developers, this shift signals that OpenAI will spend the next two years ruthlessly monetizing everything it builds. To look respectable for public investors by 2027, the era of subsidizing cheap compute is ending. Expect stricter API margins, harder upsells, and far less patience for unprofitable pet projects.

It also reshapes the AI race. If Anthropic achieves break-even enterprise operations first, or if Google continues translating its model advances directly into cloud balance sheets, OpenAI's narrative as the runaway sector leader will face a brutal reality check.