OpenAI Is Heading to Wall Street With $25 Billion in Revenue and $14 Billion in Losses — And That's the Whole Story

The most remarkable thing about OpenAI's plan to go public before the end of 2026 isn't the $25 billion in annualized revenue. It's that CEO Sam Altman and his own CFO can't agree on whether that number makes the IPO more or less dangerous. One of the most consequential internal governance standoffs in tech history is playing out quietly , and the outcome will define how the entire AI sector thinks about public markets for the next decade.

What Actually Happened

OpenAI's annualized revenue run rate surpassed $25 billion by February 2026, up from just $6 billion at the close of 2024 , a roughly 4x increase in 14 months. In the same period, the company projected it would lose $14 billion in 2026 alone. For context: OpenAI generated $13.1 billion in revenue in full-year 2025 but spent approximately $22 billion to do it. Internal projections reportedly show no path to profitability until at least 2030.

Against that backdrop, OpenAI closed a record $122 billion funding round on March 31, 2026 at a post-money valuation of $852 billion , the largest venture capital round in history. With this war chest secured, the company began what sources describe as early structural preparations for a public listing, targeting the fourth quarter of 2026. OpenAI plans to reserve a portion of IPO shares for retail investors , an unusual move for a company of this profile, signaled publicly by CFO Sarah Friar in April. The target IPO valuation: approximately $1 trillion, pricing the offering roughly 17% above the already-record private round valuation.

Why This Matters More Than People Think

The Altman-Friar disagreement is not a routine internal debate. CEO Sam Altman is pushing hard for a 2026 listing, eager to convert peak AI market sentiment into public market validation before the competitive window narrows. CFO Sarah Friar is reportedly flagging $600 billion in infrastructure commitments , data centers, compute capacity, energy contracts , as a material risk that warrants waiting until OpenAI's cost structure is better understood by management and regulators alike. That's not a conservative finance function being overcautious. That's a CFO who can see fixed obligations that don't flex with revenue and is being ignored.

The market implications extend far beyond OpenAI itself. A $1 trillion public valuation would immediately create a benchmark for every AI company's private worth. Anthropic, which entered 2026 at a $380 billion valuation, xAI at $230 billion post-Series E, and dozens of Series B and C companies have all raised on the implicit assumption that OpenAI is worth more than any of them. Once OpenAI is public and reporting quarterly financials, those private valuations face their first real market test. If OpenAI trades below its IPO price, the correction ripples through the entire sector's fundraising environment within months.

The Competitive Landscape

OpenAI's IPO faces simultaneous pressure from rivals on both the product side and the capital markets side. SpaceX is reportedly targeting a roadshow as early as June 2026, which would absorb significant institutional investor attention and bandwidth. Anthropic , which closed a $50 billion raise in May 2026 at a $900 billion valuation , has signaled an October 2026 public offering. If two or three AI giants attempt to access public markets within the same six-month window, institutional demand compression becomes an acute risk. The AI IPO market is large but not unlimited, and pricing dynamics in a simultaneous offering scenario would favor whoever moves first.

On the product side, the competitive environment has never been more intense. GPT-5.5 launched in late April 2026. Anthropic's Claude Mythos Preview is being used by select enterprises including AWS, Apple, Microsoft, Google, Cisco, and JPMorgan Chase to identify critical software vulnerabilities, including a 27-year-old bug in OpenBSD. Google's Gemini 3.1 Ultra offers a 2-million-token context window across all modalities simultaneously. xAI's Grok 5 , a 6-trillion-parameter model , is expected before Q2 2026 ends. OpenAI holds the revenue lead by a wide margin, but every major competitor is spending aggressively to close the capability gap, and public market investors will price that competitive risk into the valuation.

Hidden Insight: The $600 Billion Problem Nobody Is Talking About

The real question for OpenAI's IPO is not whether it can grow revenue. The question is whether its business model is fundamentally sound under public-market scrutiny. OpenAI's unit economics , cost per API call, gross margin by product segment, the structure of its Microsoft revenue share arrangement , have never been publicly disclosed. When they appear in an S-1 filing, the market will learn for the first time whether OpenAI's AI leadership translates into durable pricing power or whether its most important product is one competitive model release away from a price war it cannot win profitably.

The $600 billion in infrastructure commitments is the figure that should define how sophisticated investors evaluate this offering. If those commitments represent long-term energy contracts, data center construction agreements, and hardware purchase obligations, they create fixed costs that won't flex with revenue. OpenAI grew revenue 4x in 14 months , but if that growth rate normalizes in a more competitive market, the company could find itself structurally committed to spending at a rate that even $122 billion in the bank cannot sustain indefinitely. This is precisely the scenario that CFOs are trained to identify and prevent.

There is also a structural issue the IPO will force into the open: Microsoft. OpenAI's largest investor has deployed over $13 billion and holds preferential API pricing and resale rights that allow it to distribute OpenAI's capabilities while capturing significant margin. OpenAI announced the end of exclusivity beyond 2030 in early 2026, but the period between now and then represents a meaningful constraint on OpenAI's ability to fully capture its own value creation. When public investors see the Microsoft revenue share terms in the S-1, they will be making a new calculation: how much of the $25 billion in ARR does OpenAI actually keep, and what does the fully-independent gross margin look like when Microsoft is no longer a preferred partner but a market-rate customer?

What to Watch Next

The S-1 filing is the single most important disclosure event to watch. If filed before September 2026 , which the Altman timeline requires , look specifically at three numbers: gross margin by product segment, the Microsoft revenue share structure, and capital expenditure commitments over the next five years. If gross margins are below 40%, expect significant valuation repricing before the IPO prices. If the Microsoft terms are more restrictive than the market assumes, a sub-$1 trillion IPO valuation becomes likely regardless of top-line revenue growth.

Watch Anthropic's timeline as a leading indicator. If Anthropic accelerates its October 2026 listing or files an S-1 before OpenAI, it creates a competitive pressure scenario where the AI IPO window splits into two simultaneous offerings. The institutional investor base for AI IPOs is large but not infinite. Whoever prices second in that scenario likely prices lower , a dynamic that gives Altman strong incentive to move fast, and Friar equally strong grounds to insist the financials survive that scrutiny first. The tension between those two positions is the real story of this IPO.

OpenAI is heading to Wall Street with the fastest revenue growth in tech history and the deepest losses , and the uncomfortable truth is that both facts are part of the same bet on AI becoming the only thing that matters.

---