The AI IPO Rush: Decoding OpenAI's "Confidential Filing"

In the early hours of June 9, OpenAI announced it had submitted a "secret" S-1 to the SEC, kicking off its IPO process. Target valuation: up to $1 trillion. It could list as early as September. Add in SpaceX, set to list on June 12 ($1.77 trillion valuation), plus Anthropic, which has already completed its confidential filing (private valuation $965 billion) — within one month, the three AI giants are switching from the private market to the public market, a combined $4.5 trillion stepping into the open.

There's no precedent for this scale in US IPO history. In 1999, the internet-stock sprint saw 475 companies go public over the whole year — and all of them added together weren't as concentrated as these 3 are today.

What Is a "Confidential Filing"?

It's not "secret" — the SEC knows all along, but it's kept confidential from the market and the media. The company submits an S-1 to the SEC and goes through the review process, but the document isn't made public. It only becomes official 21 days before the company actually decides to list. This mechanism was established by the US JOBS Act (Jumpstart Our Business Startups Act) in 2012. The original intent was clear — to let small and mid-sized tech companies communicate privately with the SEC before an IPO, without having to lay out all their financial details, business model, and risk factors for competitors to see from the start.

Initially it was only open to "emerging growth companies" (annual revenue under $1 billion). In 2025 the SEC expanded the scope, so now large companies can use it too.

OpenAI gets three benefits from a confidential filing:

First, timing flexibility. Filing doesn't mean actually listing. The company can go through the SEC review process while watching the market. If it decides the timing is bad, it can withdraw at any point, and the market is none the wiser.

Second, competitors can't see it. An S-1 contains a huge amount of sensitive data — specific revenue mix, cost structure, customer list, sources of losses. Go the traditional public route, and the moment this is filed, rivals copy it. A confidential filing cuts most of that window out.

Third, it avoids the "drop after filing" embarrassment. With a traditional public S-1, from the time it's filed until listing, the company's share price (private valuation) and market sentiment are exposed. If something goes wrong in between, the company's image takes a hit. A confidential filing shrinks that "exposed period" down to 21 days.

Put plainly — this is leaving yourself an exit, not making a promise.

Why All Three Are Crowding Into This One Month

SpaceX, OpenAI, and Anthropic didn't coordinate this — they're fierce competitors. But they're clustering into this month because they're driven by the same market signals.

Three signals:

First, valuations have peaked. SpaceX at $1.77 trillion, OpenAI at $1 trillion, Anthropic at $965 billion. There's no one left in the private market who can absorb this scale — only a handful of funds could still lead a $10 billion+ round on their own. The only thing that can digest this scale is the public market.

Second, the AI valuation environment is just right. In the first half of 2026, global capital flowed out of tech stocks and then back in, and the market is still buying AI. This is a valuation window — wait any longer and you might run into a shift in the Fed's pace, geopolitical risk, or an AI cycle pullback, and the valuation won't hold.

Third, they're squeezing each other. After SpaceX lists on June 12, part of the public market's appetite for "giant AI companies" gets eaten up. OpenAI and Anthropic have to follow closely — they can't let SpaceX siphon off the capital alone.

That's why all three accelerated their pace from late May. This isn't corporate strategy, it's market timing forcing their hand.

That Internet Bubble Back in 1999

To understand this IPO wave, you have to go back to 1999.

That year, 475 companies went public in US markets, the vast majority of them internet companies. The Nasdaq index rose from 750 in 1995 to 5,048 in March 2000 — more than 4x in 5 years. The market was full of narratives like "new economy," "paradigm shift," and "zero marginal cost."

Then in March 2000, the bubble burst. The Nasdaq fell 75% over two and a half years, wiping out $5 trillion in market cap.

The most painful figure came later: of the 475 companies that went public in 1999, only 66 were still operating by 2018. 86% of them either went bankrupt, got acquired, or were delisted.

But note — there are a few fundamental differences between 1999 and 2026.

A Few Real Differences

I can point out a few real differences —

Difference one: profitability. The vast majority of internet companies that IPO'd in 1999 had no sustainable revenue model. Many were "losing $100 million a year, valued at $5 billion." The three companies in 2026 — SpaceX with Q1 revenue of $4.694 billion (but a $4.276 billion loss), OpenAI with Q1 revenue of $5.7 billion, Anthropic crossing $47 billion in annualized revenue. There's real revenue, just not yet profitable at scale.

Difference two: tech substance. Many 1999 IPOs were "bolting a .com onto a traditional business" — pets.com and webvan.com were both like that. The three companies in 2026 are genuinely reshaping computing / communications / logistics infrastructure — this is infrastructure-grade technology, not a marketing gimmick.

Difference three: customer base. The customers of 1999's internet companies were mostly "consumers not yet used to shopping online." Among the customers of these three in 2026 — SpaceX has NASA + the Pentagon + global telecoms, OpenAI and Anthropic have Fortune 500 enterprises. These are mature customers paying real money.

Difference four: the nature of the money. At the peak of the 1999 bubble, retail investors chasing the hype made up a huge share of IPO subscriptions. The early investors in these three in 2026 are all top-tier institutions (Sequoia, Greenoaks, Founders Fund, etc.). Retail buying comes after the listing.

Back to OpenAI's confidential filing

The signal it sends the market has two layers —

On the surface: OpenAI is getting ready to IPO, target valuation $1 trillion.

Deeper down: OpenAI choosing a confidential filing is leaving itself the option of "not going public after all." If over the next few months SpaceX's listing goes badly, AI valuations pull back, or a macro black swan shows up, it can quietly withdraw.

That posture of "keeping room to maneuver" is itself the signal most worth reading in this AI IPO wave.

Not everyone believes these valuations can hold. Including OpenAI itself.