Market thesis / Jun 26, 2026 / 5 min
OpenAI May Delay Its IPO to Protect a $1 Trillion Valuation
On June 25, the New York Times reported OpenAI is leaning toward delaying its IPO until 2027 rather than cut its $1 trillion valuation target — and SoftBank shares fell as much as 12% on June 26 as investors priced the first real crack in the AI listing parade.
OpenAI is leaning toward delaying its IPO until 2027 rather than accept a lower valuation, per a June 25 New York Times report — and CEO Sam Altman called any cut below $1 trillion a "nonstarter." The shift lands three weeks after SpaceX's record debut shed roughly a third of its value from peak and one day after Washington asked OpenAI to gate GPT-5.6 customer by customer. The AI IPO parade is still coming. The easy part is over.
What broke on June 25:
- The New York Times, citing three people involved in OpenAI's deliberations, reported the company is leaning toward a 2027 listing instead of a late-2026 debut.
- Advisers offered two paths: wait for a $1 trillion valuation in 2027, or go public sooner at a lower price. Altman rejected the discount, one person in contact with him told the Times, per Forbes.
- CFO Sarah Friar has been pushing internally for the delay, telling associates a 2027 timeline is what she has advocated, per Reuters.
- SoftBank Group shares fell as much as 12% on June 26 in Tokyo as investors recalibrated expectations for a liquidity event from one of OpenAI's largest backers.
The SpaceX warning shot:
SpaceX priced its June 12 IPO at $135 a share, surged past $225 within a week, and closed at $153 on June 25, per Forbes.
That retracement reset the calculus for every AI listing behind it. Bankers advising OpenAI warned that volatile tech markets and SpaceX's post-IPO slide could dampen retail enthusiasm for another mega-debut, the Times reported.
OpenAI filed confidentially with the SEC on June 8. Its own statement was blunt: "We have not decided on timing yet; it may be a while because there are things we want to do that are likely easier as a private company."
The filing gave OpenAI optionality. SpaceX gave it a cautionary tale.
The numbers Altman won't discount:
OpenAI's last private round valued the company at roughly $852 billion after raising $122 billion earlier in 2026, per Forbes and the Straits Times.
The gap to $1 trillion is not trivial. Forbes cited roughly $13 billion in 2025 revenue against a $21 billion net loss, with $600 billion in projected compute and hardware spending through 2030.
The Wall Street Journal reported this month that OpenAI missed recent internal revenue targets. Friar has expressed concern about the company's finances ahead of a public debut, per the Journal.
Altman is hunting new revenue streams: ads inside ChatGPT, e-commerce tie-ups with Shopify and Stripe, and cuts to money-losing ventures like the Sora video app, per Forbes.
A CFO cannot sell a prospectus built on growth stories alone. Friar needs more quarters where the numbers close the gap.
Anthropic gets first mover advantage:
Anthropic filed its confidential S-1 on June 1 — a week before OpenAI — and has been targeting an October 2026 Nasdaq debut at a reported $965 billion valuation.
The Wall Street Journal reported that bankers have told both labs whoever lists first will define how public markets price frontier AI.
If OpenAI's 2027 tilt holds, Anthropic goes first into a market that just punished SpaceX for burning cash on AI infrastructure.
Ramp's June AI Index already showed Anthropic edging OpenAI in paid U.S. business subscriptions. The IPO order may now follow the same script.
Washington adds a second gate:
The IPO timing fight is not happening in a vacuum.
On June 25, the Trump administration asked OpenAI to limit GPT-5.6 to government-approved enterprise partners before any wider release — with Washington signing off customer by customer.
That is a softer mechanism than the export-control kill switch that took Anthropic's Mythos and Fable offline. It is still a federal clearance process layered onto a product launch weeks before a company expected to price itself for public investors.
Public companies disclose regulatory risk. OpenAI is now living it in real time.
What to watch:
- Whether OpenAI confirms or walks back the 2027 timeline in the coming weeks
- Anthropic's October 2026 listing as the first live test of AI IPO appetite
- SoftBank and other OpenAI backers as delayed liquidity reprices portfolio math
- Q2 hyperscaler earnings for evidence AI revenue is catching up to AI capex
- OpenAI's draft S-1 disclosures on losses, infrastructure commitments, and copyright litigation when the filing goes public
Convina's view: OpenAI spent June filing for an IPO, missing revenue targets, accepting federal customer vetting on GPT-5.6, and watching its CFO argue the company is not ready for Wall Street. Altman's refusal to accept anything below $1 trillion is not vanity — it is the only way to justify $600 billion in infrastructure bets to public investors. But SpaceX just proved mega-debuts can crater fast, and SoftBank's June 26 selloff shows the collateral damage is already spreading. The smartest move is not to rush the listing. It is to let Anthropic go first and see if the market still believes the story.