OpenAI confidentially filed its S-1 registration statement in May, formally beginning the public-listing process -- and is now reportedly reconsidering the timeline entirely. Reuters reported in late June that OpenAI is weighing a delay to 2027, with The New York Times and Forbes separately reporting similar hesitation, citing volatile public tech markets and the underwhelming post-IPO trading pattern of SpaceX, which debuted just weeks earlier, as a cautionary comparable.
The financial picture explains the tension. OpenAI says it's generating roughly $2 billion in revenue per month, a growth trajectory the company describes as four times faster than Alphabet or Meta achieved at comparable stages -- genuinely exceptional top-line growth by any historical software-company standard. But the company also reportedly projects around $14 billion in losses for 2026 alone, with internal estimates not expecting profitability until roughly 2030. That combination -- extraordinary growth alongside years of continued heavy losses -- is a much harder story to tell public-market investors than private ones already bought into the long-term AI thesis.
SpaceX's post-IPO trading is the immediate cautionary tale informing OpenAI's hesitation: shares spiked to an all-time high just two trading days after its June 12 debut, then saw the initial rally fade and selling pressure intensify through the following week, before stabilizing and regaining strength heading into this week's Nasdaq-100 inclusion. If a company with SpaceX's brand recognition and Elon Musk's following can see that much immediate volatility, OpenAI's board has reason to worry about how a similarly loss-making, though faster-growing, AI company would trade.
“But the company also reportedly projects around $14 billion in losses for 2026 alone, with internal estimates not expecting profitability until roughly 2030.”
Anthropic's own choice is instructive by contrast: rather than accelerating toward a public listing, Anthropic raised $65 billion in a private round in May at a valuation near $1 trillion, opting to let private investors continue absorbing its losses rather than testing public-market appetite. That suggests private capital remains abundant and willing to fund frontier-lab cash burn at valuations rivaling or exceeding what public markets might currently support, reducing the urgency for either OpenAI or Anthropic to rush a listing.
Compared to the broader 2026 tech IPO calendar -- Bending Spoons' volatile but ultimately successful debut, and a healthy overall pipeline running roughly 8% ahead of 2025's IPO count -- the hesitation from OpenAI specifically looks less like a sign of a broken IPO market and more like company-specific caution about how a truly enormous, deeply loss-making AI lab would be received relative to smaller, profitable software companies.
For growth and pre-IPO investors, a 2027 OpenAI listing (if it materializes) means continued illiquidity for existing shareholders and employees for at least another year, likely increasing demand for secondary-market transactions and tender offers in the interim to provide liquidity without a full public listing.
The bear case against reading too much into this: IPO timing speculation is notoriously unreliable, and companies frequently file confidentially and then wait well over a year before actually pricing -- OpenAI delaying to 2027 wouldn't be unusual by historical S-1-to-pricing timelines, even without SpaceX's choppy debut as a factor.
What to watch: whether OpenAI makes any formal statement on IPO timing, how SpaceX's stock performs over the next several months as a live comparable, and whether Anthropic's private-capital strategy (versus OpenAI's public filing) becomes the more common path for frontier labs delaying public listings.