Nobody is ready for the AI IPO wave

Sometime in late 2026 or early 2027, OpenAI and Anthropic will almost certainly go public. Both are approaching or already at trillion-dollar private valuations. Both are burning through capital at a rate that makes continued private fundraising unsustainable. And both are about to expose their financials to a public market that has never seen anything quite like this. This isn't just another tech IPO cycle. This is two companies, built on the same foundational technology, racing to list at valuations that would place them among the ten most valuable companies on the planet, before either has turned a profit. The public markets are about to get a reality check on what these companies actually earn versus what they promise.

The numbers so far

OpenAI raised $122 billion in March 2026 at an $852 billion valuation. It generates roughly $2 billion in revenue per month, putting its annualized run rate above $25 billion. That sounds enormous, and it is. But OpenAI's own internal forecasts project a $14 billion loss in 2026. Revenue is growing fast, but compute costs are growing faster. Then came the April bombshell. The Wall Street Journal reported that OpenAI had missed several internal revenue targets and failed to hit its goal of one billion weekly active users for ChatGPT by the end of 2025. CFO Sarah Friar reportedly warned colleagues that if revenue growth doesn't accelerate, the company may struggle to fund its future computing contracts. OpenAI pushed back on the report, but the damage was done. Shares of key investors like SoftBank, Oracle, and Nvidia all dropped on the news. Anthropic's trajectory looks different, and in some ways more impressive. The company hit $30 billion in annualized revenue in April 2026, up from $9 billion at the end of 2025 and just $1 billion in December 2024. That kind of growth has no precedent in enterprise software. More than 1,000 business customers now spend over $1 million per year on Claude. Anthropic raised $30 billion at a $380 billion valuation in February, and by late April, reports emerged that it was weighing a new round at $850 to $900 billion. Both companies are now in IPO preparation. Anthropic tapped law firm Wilson Sonsini in late 2025 to begin work on what could be one of the largest public offerings ever. OpenAI has been expanding its finance team and building out investor relations. The question isn't whether they'll go public. It's what happens when they do.

The margin problem

Revenue is the easy story. Margins are the hard one. Traditional SaaS companies go public with gross margins of 70 to 85 percent. Server costs are low. Each additional customer costs almost nothing to serve. That's the beauty of software economics, and it's what public market investors have been trained to expect. AI companies break this model. Every query, every API call, every "generate" button triggers expensive GPU compute. Anthropic spent $6.8 billion on compute in 2025 alone, split between model training and inference. OpenAI's compute commitments are even larger, with the $500 billion Stargate Project representing the scale of infrastructure these models demand. The result is that revenue can grow exponentially while losses grow right alongside it. OpenAI's $25 billion in revenue sounds like a business that should be wildly profitable. But when your compute bill might exceed your revenue, as OpenAI's own CFO has reportedly warned, the unit economics look nothing like the SaaS playbook that public investors understand. This is the core tension that will define how these IPOs are received. Investors will need to decide whether AI margins will eventually look like software margins, or whether the compute-intensive nature of the business creates a fundamentally different, and less attractive, economic structure.

When hype meets the public market

History offers some uncomfortable parallels. Facebook went public in May 2012 at $38 per share, valuing the company at over $104 billion. It was the most anticipated tech IPO of its era. Within four months, the stock had dropped more than 50 percent, hitting $17.73 by early September. Morgan Stanley had to aggressively buy shares on day one just to keep the price from falling below the offering price. Retail investors who bought in at the IPO were underwater for over a year. Facebook eventually recovered and then some. A $100 investment at IPO is now worth over $1,800. But the recovery took years and required Facebook to prove it could monetize mobile, which wasn't obvious at the time of listing. Uber's 2019 IPO tells a different story. The company went public at $45 per share, dropped 7.6 percent on its first day, and spent years trading below its IPO price. WeWork's attempted IPO that same year was even worse, collapsing entirely when public market scrutiny revealed a business that couldn't justify its private valuation. The pattern is consistent: when companies go public at peak hype, with valuations priced for perfection, the margin for error is zero. Any stumble, whether it's slowing growth, rising costs, or competitive pressure, gets punished severely. OpenAI and Anthropic will face this dynamic at an unprecedented scale. A trillion-dollar IPO valuation assumes not just continued growth, but continued dominance in a market that is evolving faster than almost any in history.

Who actually benefits

IPOs are often framed as a moment of validation for a company. In reality, they're a liquidity event, and the question of who benefits depends entirely on when you got in. Early investors and employees at OpenAI and Anthropic stand to gain enormously. Venture firms that invested at single-digit billion valuations will see returns that define the decade. Employees with stock options will see life-changing payouts. But for late-stage investors who came in at $380 billion or $852 billion? The math is much harder. If Anthropic IPOs at a $600 to $900 billion market cap, the upside for its most recent investors is modest at best. If OpenAI lists at or near its last private round valuation, the latest investors are essentially flat. And for retail investors, the ones who can finally buy in on IPO day? They're getting the last seat at the table. The explosive growth phase has already been captured by private capital. What's left is the harder work of proving the business model at scale, in public, with quarterly earnings calls and analyst scrutiny. Microsoft, which owns roughly 27 percent of OpenAI, will be a major beneficiary regardless. Amazon, with its massive investment in Anthropic, is similarly positioned. For investors who want exposure to these IPOs without the single-stock risk, the public companies that already own significant stakes may be the more rational play.

What changes when the curtain lifts

The most underappreciated consequence of these IPOs has nothing to do with stock prices. It's about transparency. Private companies enjoy extraordinary opacity. OpenAI and Anthropic can announce revenue milestones on their own terms, share growth metrics selectively, and keep their cost structures entirely private. That privilege disappears the moment they file an S-1. Public markets will demand answers to questions that have so far been optional. What are the actual gross margins after compute costs? How much revenue comes from a handful of large enterprise contracts versus a diversified customer base? What's the customer churn rate? How sustainable is the growth without continued massive capital raises? There will also be scrutiny on training data practices, safety investments, and competitive dynamics. Anthropic has built its brand around AI safety, but public investors will want to know exactly how much that costs and whether it creates a competitive moat or just a line item. OpenAI's restructuring from a nonprofit to a for-profit public benefit corporation will face questions about governance and mission alignment. For the broader AI industry, these filings will be the first real window into the economics of frontier AI development. Every startup, every investor, and every competitor will study those numbers to understand whether building large language models is actually a good business, or just an expensive one.

What this means for the ecosystem

The ripple effects of these IPOs will extend far beyond OpenAI and Anthropic. If the IPOs go well, with strong first-day pops and sustained trading above the offering price, it will validate the entire AI investment thesis. Venture capital will pour even more aggressively into AI startups. Enterprise spending on AI will accelerate as procurement teams point to public market performance as evidence that AI is here to stay. If the IPOs stumble, with flat or declining stock prices in the months after listing, it could trigger a broader correction across the AI ecosystem. Late-stage AI startups will find fundraising harder. Public AI stocks that have run up on sentiment could face a reckoning. The narrative could shift from "AI is transforming everything" to "AI is expensive and unproven," even if the underlying technology continues to improve. The most likely outcome is somewhere in between. These are real companies with real revenue and real products that millions of people and businesses use every day. But they're also companies burning billions in cash, operating in a fiercely competitive market, and valued at levels that assume near-perfection for years to come.

The bottom line

The AI IPO wave is coming, and nobody is fully prepared for it. Not the companies, which will face a level of scrutiny they've never experienced. Not the public markets, which have no real framework for valuing businesses with this combination of explosive growth and massive losses. And not the retail investors who will finally get their chance to own a piece of the AI revolution, likely at the most expensive price it's ever been offered. The question isn't whether OpenAI and Anthropic are important companies. They obviously are. The question is whether importance and investment return are the same thing, and history suggests they often aren't.