ChatGPT’s Growth Is Stalling, and That’s a Real Problem for OpenAI’s IPO

ChatGPT had a hell of a run. For a while it felt like everyone and their grandmother was prompting it for recipes, emails, and bad poetry. But the honeymoon phase is clearly over.

Sensor Tower just dropped some numbers that should worry OpenAI. In April, uninstalls of the ChatGPT app jumped 132% year-over-year. That’s bad enough. But March was worse — a staggering 413% increase in uninstalls compared to the same month last year. That spike lines up neatly with OpenAI’s controversial Pentagon deal announced in February. Users who were already uneasy about military applications apparently decided to vote with their thumbs.

Now, let’s be fair — ChatGPT is still adding users overall. The absolute numbers are huge. But the growth rate is cratering. Sensor Tower reports that monthly active users grew 168% in January, but only 78% in April. That’s a steep deceleration in just three months. If you’re plotting a hockey-stick growth curve for an IPO prospectus, this is not the shape you want to see.

OpenAI is reportedly eyeing a valuation north of $80 billion for its upcoming IPO. Investors love a story about unstoppable adoption. They hate seeing competitors like Anthropic’s Claude, Google’s Gemini, and a dozen open-source alternatives nipping at the leader’s heels. When users can switch to a rival with zero friction, retention becomes everything.

The Pentagon deal was a calculated risk. OpenAI wanted to show it could land big enterprise contracts, especially in defense. But a chunk of the consumer base — the very people who made ChatGPT a cultural phenomenon — clearly doesn’t love that direction. The uninstall spike is a direct signal.

I’ve been saying for a while that the AI chatbot market was due for a shakeout. First-mover advantage only lasts as long as you keep delivering something the others can’t match. Right now, ChatGPT’s core product hasn’t changed dramatically in months. The GPT Store was a flop by any measure. The voice mode is neat but not a game-changer. Meanwhile, Claude is writing better long-form content, Gemini is deeply integrated into Google’s ecosystem, and open-source models are catching up fast.

OpenAI still has the brand and the installed base. But slowing growth plus rising churn is a dangerous combination. If the IPO happens in the next year, the underwriters are going to have to work hard to spin these numbers.

I don’t think ChatGPT is dying. Far from it. But the days of effortless, exponential growth are over. Now we see whether OpenAI can actually retain users and justify that massive valuation. The Pentagon deal bought them a customer. It might have cost them a lot more.