OpenAI and Anthropic IPO watch: confirmed versus speculation

The two most valuable private companies in AI both took the same formal step toward public markets within a week of each other this June, and the coverage since has blurred a line that matters enormously: what the companies have actually said versus what everyone else is saying about them. This piece keeps that line sharp, because if you are planning around these IPOs, as an investor, a customer, or a competitor, the difference between confirmed and speculated is the whole game.

What is actually confirmed

Anthropic said on June 1 that it confidentially submitted a draft registration statement on Form S-1 to the SEC for a proposed initial public offering. Its own words are carefully hedged: the filing “gives us the option to go public after the SEC completes its review,” with shares and pricing not yet set and everything dependent on market conditions.

OpenAI followed on June 8 with an unusually candid confirmation: “We recently submitted a confidential S-1. We expect it to leak so we’re just announcing it. 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.” That is the entire confirmed record. Two confidential draft filings, two explicit statements that nothing is decided.

What is reported but not confirmed

Outlets including CNBC have reported Goldman Sachs and Morgan Stanley working with OpenAI, a potential fall 2026 window for both companies, an October Nasdaq listing target for Anthropic near a $965 billion valuation, and OpenAI ranges from $730 billion, its last funding round’s pre-money mark, up past $1 trillion. Late-June reporting in the New York Times and Forbes, citing unnamed sources, suggested OpenAI may lean toward delaying into 2027. None of this comes from the companies, and the delay reports directly contradict the fall-window reports, which tells you how soft the sourcing still is.

Why the confidential step matters anyway

A confidential S-1 is cheap optionality, but it is not nothing: it forces audited financials, SEC review, and internal readiness that private companies otherwise defer. It also lands amid exactly the market questions we laid out in our AI bubble evidence review: an IPO is the first moment public markets get to price what private rounds have asserted, at revenue multiples that currently run 20 to 40 times. And as we noted covering Altman and Amodei’s softened public messaging, both CEOs began repositioning their public statements in the same weeks these filings landed, a coincidence worth exactly as much skepticism as outlets like Fortune have given it.

For business buyers, the practical angle: public AI labs would mean quarterly disclosure of the unit economics we can only estimate today, the margins question at the center of our margin scoreboard analysis, and real pricing pressure transparency for anyone negotiating enterprise AI contracts.

What to watch

The next hard signals, in order: a public S-1 flip (the filings becoming visible, with real financials), a roadshow announcement, and pricing. Anything before those three is atmosphere. If either company’s public filing reveals financials meaningfully weaker than the reported ARR figures, the repricing conversation stops being theoretical, for the whole sector at once.

Frequently asked questions