OpenAI IPO: Microsoft Is Both Partner and Risk

The Company That Could Be Worth a Trillion Dollars Just Told Investors Its Biggest Partner Might Also Be Its Biggest Problem

On March 23, 2026, OpenAI did something that almost no startup preparing for a blockbuster IPO ever does voluntarily: it told prospective investors, clearly and in writing, that the company most responsible for its survival could also be the company that destroys it. The document circulated to investors — which reads in every meaningful way like an IPO prospectus without the SEC filing stamp — contains a risk factors section that names Microsoft directly. Not obliquely. Not buried in footnote seventeen. Right there in the main text, OpenAI wrote that Microsoft supplies a "substantial portion of our financing and compute," and that if the partnership were modified or terminated, the consequences for OpenAI's "business, prospects, operating results and financial condition could be adversely affected." This is a company targeting a public market valuation of $830 billion to $1 trillion telling investors its foundation might be borrowed.

To understand why this disclosure landed with the force it did on March 25, 2026, you need to understand the timeline of the relationship it describes. Microsoft has invested $13 billion in OpenAI since 2019. Its 27 percent diluted stake in OpenAI Group PBC is currently valued at approximately $135 billion. The partnership built ChatGPT on Azure infrastructure, gave Microsoft the AI capabilities that Satya Nadella used to redefine the company's competitive position against Google, and turned what was a nonprofit research lab into the most valuable private company in the history of technology. And now, in the same week that OpenAI is circulating what is functionally its IPO prospectus, that partnership is showing its first visible cracks.

The AWS Deal That Changed Everything

The fracture did not start with the investor document. It started on February 27, 2026, when OpenAI signed a $50 billion deal with Amazon Web Services, granting AWS exclusive hosting rights for Frontier, OpenAI's AI agent platform. Microsoft had invested $13 billion in OpenAI specifically to secure cloud exclusivity for its Azure platform under a binding contract. That contract, in OpenAI's own language from its restructuring documents, was a foundational element of the commercial arrangement. The AWS deal broke it. According to reporting from The Globe and Mail, Microsoft is now considering legal action over the alleged breach of contract. The company that is listed as OpenAI's most critical infrastructure partner is evaluating whether to sue it.

Meanwhile, Microsoft has spent the months since November 2025 building a Superintelligence team of its own — recruiting researchers from OpenAI, Google DeepMind, and other competitors. The move signals that Microsoft is reducing its own dependence on the partnership even as it publicly maintains that the relationship remains strong. Both sides are hedging. Both sides are building exits. And in the middle of this quiet strategic decoupling, OpenAI is trying to convince public market investors to value it at close to a trillion dollars.

The Numbers Behind the Headline

OpenAI's investor document arrives at a moment of spectacular commercial success and equally spectacular financial loss. ChatGPT now has 900 million weekly active users. The company generated $13.1 billion in revenue in 2025. It was valued at $730 billion in its most recent funding round, which closed at $110 billion from investors including Amazon, Nvidia, and SoftBank. By any measure of product adoption, OpenAI is the dominant generative AI company on earth.

The financial projections underneath that success are considerably darker. OpenAI is projected to lose $14 billion in 2026 alone. Cumulative cash burn is expected to reach $115 billion through 2029. HSBC projects a $207 billion funding shortfall by 2030. The company has committed $250 billion to Microsoft Azure, $300 billion to Oracle Cloud, and $38 billion to AWS — infrastructure commitments that total approximately $1.4 trillion through 2033. Internal projections show profitability arriving in 2029. OpenAI's public statements say profitability is not expected before 2030. Even the company's optimistic and pessimistic scenarios disagree by a year on when the losses stop.

The market share picture adds another dimension of complexity. ChatGPT's web traffic share fell from 86.7 percent to 64.5 percent in twelve months as Google Gemini surged to 21.5 percent. The product that built OpenAI's dominant position is being eroded in real time. That erosion, combined with the Microsoft dependency disclosure, the AWS contract breach, and the loss projections, represents a risk profile that is genuinely unusual for a company seeking a trillion-dollar valuation.

Elon Musk and the $134 Billion Trial Coming in April

The Microsoft situation is not the only legal and financial exposure that OpenAI's investor document is required to disclose. The document also flags the Musk v. OpenAI trial, set to begin April 27, 2026, in Oakland, California — exactly one month from the date of this writing. Musk, who co-founded OpenAI in 2015 with a $38 million seed donation and claims to have provided roughly 60 percent of the nonprofit's early funding, is seeking between $79 billion and $134 billion in damages from OpenAI and Microsoft combined.

The damages calculation comes from financial economist C. Paul Wazzan, whose analysis concludes that OpenAI captured between $65.5 billion and $109.4 billion in wrongful gains from Musk's contributions, while Microsoft captured an additional $13.3 billion to $25.1 billion. The legal theory is that OpenAI's conversion from a nonprofit research lab to a for-profit Public Benefit Corporation — which Musk claims happened without his consent and contrary to the founding agreement — constitutes fraud. OpenAI calls the lawsuit "baseless" and part of what it describes as a "harassment campaign." A federal judge has already rejected a final bid by OpenAI and Microsoft to avoid a jury trial. The case is going forward whether or not either company wants it to.

The timing is extraordinary. OpenAI is circulating IPO prospectus materials in March 2026 while a jury trial seeking damages that could exceed a hundred billion dollars is scheduled for April 2026. If Musk wins even a fraction of what he is asking, the financial consequences would be catastrophic for a company already projecting $14 billion in losses for the year. If he loses, the overhang disappears. But until a verdict is rendered, every investor reading OpenAI's pre-IPO document has to price the uncertainty into their valuation.

TSMC, Chips, and the Supply Chain That Cannot Be Controlled

Beyond Microsoft and Musk, OpenAI's investor document identifies a third category of existential risk: semiconductor supply chains. The company is heavily dependent on chips manufactured by TSMC, the Taiwan Semiconductor Manufacturing Company that produces the overwhelming majority of the world's most advanced processors. TSMC's geographic position — its primary fabrication facilities are in Taiwan — places it at the intersection of the US-China geopolitical competition that has defined technology policy for the past several years. Any disruption to TSMC's operations, whether from geopolitical conflict, natural disaster, or production failure, would directly constrain OpenAI's ability to train and deploy its models. For a company whose entire product is software that requires enormous quantities of specialized compute, that dependency is not theoretical. It is structural.

The TSMC risk factor appears in the same document as the Microsoft dependency and the Musk lawsuit, which means public market investors who eventually read OpenAI's formal S-1 filing will encounter a company that has, with unusual candor, outlined three categories of risk that each individually could be company-defining. The combination of all three in a single IPO filing is, in the recent history of major technology listings, genuinely unprecedented.

What the IPO Actually Means for the AI Industry

OpenAI's IPO, when it arrives — and most analysts now expect a formal S-1 filing with the SEC sometime in mid to late 2026 — will be more than a corporate finance event. It will be a referendum on whether the entire investment thesis behind generative AI holds up under public market scrutiny. Hundreds of billions of dollars have been committed by Microsoft, Nvidia, Amazon, and Oracle to AI infrastructure, much of it flowing directly through OpenAI. If public investors accept a valuation of $830 billion to $1 trillion for a company losing $14 billion per year with a disputed foundational partnership and a hundred-billion-dollar lawsuit set for trial in four weeks, it validates the largest capital deployment cycle in technology history. If they don't, the shock propagates through every AI investment made on the assumption that this market was rational.

For the global technology industry in March 2026, OpenAI's pre-IPO investor document is not just a corporate disclosure. It is the most honest public accounting of how fragile the AI boom's infrastructure actually is — written by the company at the center of it, for the investors being asked to price it. Sam Altman and his team did not have to be this candid. The fact that they were suggests they believe the story survives the scrutiny. Whether the public markets agree is the question that will define the second half of 2026.