No OpenAI S-1 Filed as Banks Gear Up for $1 Trillion IPO

Goldman and Morgan Stanley have a credibility problem they can't yet see.

There is no OpenAI S-1 — the SEC's formal IPO registration statement — on the Commission's public EDGAR database as of Friday, May 23. When that filing surfaces — expected ahead of a September Nasdaq listing, per multiple reports citing people familiar with the matter — it will be the first time the company has publicly disclosed revenue figures, burn rate, and risk factors in a regulatory document. Goldman Sachs and Morgan Stanley are leading what would be the largest tech IPO in history, a potential $1 trillion listing for a company that has never filed one before. The bankers know what they're selling. What they don't know is what they'll have to say about the record that came before it.

A federal jury in Oakland spent less than two hours last Monday and returned a unanimous verdict: Elon Musk filed his $150 billion lawsuit too late. The statute of limitations — California's three-year limit for fraud claims — cut the case off before the jury could weigh whether Sam Altman had actually deceived him, per Reuters's courtroom coverage. What the jury never reached: the part where five witnesses called Altman a liar under oath during a three-week trial, per The Guardian's trial coverage. Their testimony is in a federal docket. S-1 risk disclosures routinely reference material litigation. Whether those five witnesses appear in the next document is the question Goldman and Morgan Stanley need answered before they price the largest AI IPO in history.

The court record is real. The credibility gap is real. The filing that would answer how OpenAI addresses it does not yet exist.

OpenAI has reported raising $122 billion this year at an $852 billion valuation, with up to $1.4 trillion in data center commitments over the next decade, per MSN's reporting. Those figures reflect private market valuations — the gap between what private investors paid and what public markets will accept is the central risk the S-1 will have to bridge. Wedbush analyst Dan Ives called the courtroom outcome "a huge win for Altman and OpenAI despite the scrapes and bruises on Altman's persona and leadership," per Reuters. That framing — win, with bruises — is the one the banks will use. Whether pension funds and sovereign wealth funds, who read the same federal docket the banks do, find it credible is another question.

The comparison that matters: SpaceX filed its S-1 with the SEC on May 20, targeting a June 11 pricing on Nasdaq, per its registration statement on SEC EDGAR. SpaceX's filing contains verified financials, audited risk disclosures, and a founder record that Musk controls through 85% voting power. OpenAI's filing will contain none of those things — no full financials, no independent audit, no prior public track record — and a federal record with five witnesses who called the CEO a liar. Institutional investors sizing up the two offerings at the same moment will notice the difference.

There is a counterargument worth stating plainly. Altman's allies will argue the jury didn't reject his character — it rejected Musk's legal theory. The statute of limitations is a procedural bar, not a verdict on truth-telling. Five witnesses with varying motives testified to a board dispute that predates the for-profit restructuring. The jury reached the right result for the wrong reason, and the S-1 risk factors will say nothing about witness credibility. This argument is coherent. It is also the one the banks will have to make in the same document where they ask pension funds to trust the world's most valuable private AI company with their capital.

What the jury did not do, however, was find that Altman told the truth.

Five witnesses called Altman a liar under oath during the three-week trial, according to The Guardian's trial coverage and other outlets with reporters in the courtroom. Their testimony is now part of the permanent federal record. Among them: former OpenAI employees who described internal disagreements Altman had denied. A journal entry by Altman himself was cited as inconsistent with his in-court account. Musk's legal team argued the entire OpenAI restructuring was a scheme to benefit Microsoft and private investors at the expense of the nonprofit's stated mission. Satya Nadella, Microsoft's CEO, testified he never got a clear explanation from the board for why Altman was fired in 2023. Ilya Sutskever, OpenAI's former chief scientist, told the board Altman exhibited a consistent pattern of lying and pitting executives against each other, per Guardian coverage. Mira Murati testified Altman had a habit of telling one person the opposite of what he told another.

None of that went to the jury on the merits. The statute of limitations cut the case off before the substance could be weighed, per Reuters. Altman's lawyers argued he knew about OpenAI's growth plans years earlier and slept on his rights. The jury agreed. Case over.

The verdict removes one overhang. It creates another. Goldman and Morgan Stanley will have to decide whether to characterize the outcome as a vindication — which it technically is, on the narrow question of when Musk sued — or as a verdict that sidestepped the hard question entirely. That decision sits in a filing that doesn't exist yet. When it appears, it will be the most read document in Silicon Valley.

The courtroom win was never the story. The filing that hasn't arrived is.

This story will be updated when OpenAI files its S-1 with the SEC.