The New York Times reports OpenAI is “leaning toward” delaying its IPO to 2027. Choppy markets. Concerns about retail investor appetite. Advisors recommending either wait for $1 trillion or lower the valuation. The company that spent this week announcing Jalapeño chips, $100B ad ambitions, and GPT-5.6 just blinked on the biggest bet of all.
The IPO Dilemma
2027
New target — pushed from 2026
$1T
Target valuation advisors recommend
$850B
Last private valuation
$3.7B
Q1 operating loss — still burning
What the NYT Reports
Three factors driving the delay:
Choppy markets. SpaceX IPO’d at $135, rocketed to $225+, then crashed back to ~$153. The mega-IPO playbook just showed its risk.
Retail investor skepticism. OpenAI is worried it “may not find much enthusiasm from retail investors” — the lifeblood of a consumer-facing IPO.
Valuation gap. Advisors say: wait until 2027 to IPO at $1 trillion, or lower the valuation for a quicker listing. Neither is great — one delays cash, the other signals weakness.
Internally, CFO Sarah Friar reportedly urged waiting until 2027, citing the $3.7B quarterly burn, compute infrastructure commitments, and the burden of public reporting. CEO Sam Altman favored a quicker timeline.
The Structural Read
THE SPACEX IPO SCARED EVERYONE
SpaceX’s stock collapse from $225 to $153 after its IPO is the cautionary tale. If a $2T company with real revenue (rockets + Starlink + $27B in AI compute) can lose 30% post-IPO, what happens to a company burning $3.7B/quarter? OpenAI saw the SpaceX trajectory and reconsidered.
THE BURN RATE IS THE REAL PROBLEM
$3.7B quarterly loss. $25B revenue. Negative margins. Jalapeño promises 50% inference savings — but it won’t deploy until late 2026. The ad business targets $100B by 2030 — but it’s barely started. Public markets want profitability trajectory, not narrative. OpenAI doesn’t have one yet.
The Bottom Line
OpenAI spent this week building the most aggressive IPO narrative in tech history — custom chips, $100B ad target, enterprise model launch, cybersecurity platform. Then the market sent a message: not yet. SpaceX’s post-IPO decline spooked everyone. Retail investor skepticism is real. And a company burning $3.7B/quarter can’t afford a weak debut. The delay to 2027 is rational — but it gives Anthropic, which has better margins and better models, a chance to IPO first. The most consequential race in AI might not be about models at all. It might be about who goes public first — and who the market trusts more.
Sources: New York Times, BeInCrypto — June 25, 2026