OpenAI’s IPO story is still huge, but it’s no longer as clean as it looked when ChatGPT had the category almost to itself.
OpenAI filed confidential IPO paperwork on June 8, and the timing tells you plenty. The company behind ChatGPT is trying to move toward public markets while its lead in consumer AI is still enormous, but also visibly less lonely. You can still call ChatGPT the default AI assistant. You just can’t call it the only serious one anymore.
According to Sensor Tower data reported by EFE and El Pais, ChatGPT reached 1 billion monthly active users in May 2026, making it the fastest app to reach that mark. That’s a brutal number for any rival to stare at. Gemini was second in the same Sensor Tower ranking with 472 million monthly active users, while Claude had 56 million, behind Doubao, Dola, DeepSeek and Meta AI. So no, ChatGPT hasn’t lost the race. Anyone saying that is getting carried away.
But public investors don’t only pay for where a company is today. They pay for where the curve is pointing. Sensor Tower said ChatGPT’s monthly active users were up 62% year over year in the second quarter, while Claude grew 640% and Meta AI grew 973% after Meta pushed the app through its own distribution machine. Those are not equal businesses, and they’re not equally monetized. Still, they show the part of the story OpenAI would rather not have to explain on a roadshow: the category is broadening faster than ChatGPT’s brand advantage is hardening.
That’s the real issue. In 2023, if you wanted a consumer AI assistant that felt like the future, you opened ChatGPT. By mid-2026, you’ve got Gemini inside Google’s world, Claude with a stronger reputation among many professional users, Meta AI riding WhatsApp, Instagram and Facebook, and smaller apps trying to own narrower jobs. The switching cost is not zero, but it’s low enough that habit alone won’t protect OpenAI forever.
Google is the most obvious problem because distribution is not a feature you can bolt on later. Gemini sits beside Gmail, Workspace, Android, Chrome and search. OpenAI can build better models, better products and better pricing, but it can’t wake up tomorrow with Google’s surface area. If you run a startup, you already know how this works. The best product doesn’t always beat the product already sitting in the customer’s workflow.
Claude is a different kind of pressure. Its user base is much smaller, but Sensor Tower’s reported growth shows Anthropic has become a real destination rather than a safety-brand footnote. That matters for OpenAI’s enterprise pitch because corporate buyers are not loyal in the same way fans of a consumer app are loyal. They test, compare, negotiate and switch when the numbers or policies work better. OpenAI still has Microsoft, the ChatGPT brand and a huge developer mindshare advantage. Anthropic has made the argument harder.
Then there is the IPO filing itself. The Guardian reported that OpenAI confirmed it had submitted a confidential S-1 to the SEC on June 8, while saying it had not decided on timing and may stay private for a while. The same report put the expected valuation above $850 billion and cited Wall Street Journal reporting that OpenAI closed a $122 billion funding round in March at about an $852 billion valuation. That’s not a normal software multiple. That’s a bet that OpenAI becomes one of the core operating systems of the AI economy.
Frankly, that bet can still work. ChatGPT’s scale is not a rounding error. OpenAI reported 900 million weekly active users by late February, and Sensor Tower’s May figure puts the monthly audience at a level most consumer technology companies never touch. If even a modest slice of those users turn into paid subscribers, enterprise seats or commerce and advertising revenue, the business can be enormous.
The catch is that OpenAI is asking future public investors to believe two things at once: that AI assistants will become a gigantic market, and that OpenAI will keep enough of the profit pool to justify a valuation near the top of global technology. The first claim looks stronger every month. The second is where the argument gets messy.
Regulators are adding to that mess. The Wall Street Journal, cited in recent coverage from Tom’s Hardware, reported that a coalition of 42 state attorneys general served OpenAI with a subpoena on June 12 seeking documents on advertising, data practices, minors, seniors, health information, user engagement and model behavior. OpenAI said it takes the concerns seriously and will engage with the offices. A subpoena is not a finding of wrongdoing, but if you’re preparing investors for an IPO, it’s not the kind of four-day follow-up you want after announcing a confidential filing.
None of this makes OpenAI weak. It makes the story more adult. The company is no longer selling the market a clean tale about one runaway chatbot becoming the obvious interface for everything. It’s selling a harder proposition: that a heavily scrutinized, capital-hungry AI company can keep growing into an extraordinary valuation while Google, Anthropic, Meta and others attack different parts of the same market.
That’s still a story investors may buy. But it needs facts, not aura. ChatGPT has the billion-user headline. Gemini has the distribution. Claude has momentum with serious users. OpenAI has the brand, the scale and the biggest IPO spotlight in technology. Now it has to prove those are enough.
Also read: Andreessen Horowitz bets $9 million that AI reliability is a category, not a feature • Plaud reached $250 million in recurring revenue without a single venture dollar and is now targeting $500 million in 2026 sales • Mobileye bets its sensor stack can make the leap from supplier to robotaxi operator
