OpenAI is playing a high-stakes game of Schrödinger's exit: it wants to go public, but it also wants to stay private if things get too interesting. According to leaked Slack messages and confidential SEC filings, CEO Sam Altman told employees he expects an IPO within the next year. Yet he's leaving the door open to keep the company private if recursive self-improvement in AI starts to accelerate. That's not indecision—that's a hedge against a future nobody can predict.
The timing is everything. OpenAI is preparing a new model called GPT-5.6, an iterative leap rather than a full-generation jump. At the same time, the company is rolling out a major ChatGPT update designed to turn it into a 'superapp.' These product moves aren't just about impressing users—they're about showing Wall Street that the revenue pipeline is real. And the revenue is real: OpenAI pulled in more than $13 billion last year. But here's the kicker—the company expects to spend $115 billion over the next four years. That's a lot of AI-generated cat memes to sell to cover those costs.
The IPO narrative gets even messier when you look at Altman's recent backtracking. He's walking back earlier grand predictions about superintelligence and AI's impact on labor. That's a problem. Investors love a visionary, but they hate one who changes the story mid-pitch. Meanwhile, rival Anthropic is also preparing for a $1 trillion valuation IPO, and its CEO Dario Amodei is similarly softening his own apocalyptic forecasts. The two labs are locked in a race that isn't just about who reaches Wall Street first—it's about who can convince the public that their AI won't eat our jobs before the bell rings.
The numbers are dizzying. A tender offer priced shares at $687 each, and Fortune pegs both OpenAI and Anthropic at roughly $1 trillion valuations. But a $1 trillion price tag on a company that burns cash at this rate? That takes guts. Or desperation. Or a new model that makes GPT-5.5 look like a pocket calculator.
What should you care? If OpenAI IPOs, you might get to buy a piece of the AI revolution. But if Altman decides to stay private because his AI is improving too fast, you'll be left watching from the sidelines. Either way, the next year will define whether OpenAI is the next Google or the next cautionary tale.
One thing is certain: GPT-5.6 is coming, the superapp is coming, and the IPO clock is ticking. But the most dangerous variable isn't a model or a filing—it's Altman's own changing narrative. Can he sell a future he's no longer sure he believes in? That's the real question the market will answer.