ChatGPT growth slows as app deletions climb, raising risks for OpenAI's IPO

OpenAI is facing an unwelcome signal for its main consumer product: ChatGPT's growth is slowing while the number of app deletions climbs noticeably. With a possible IPO under discussion, an ordinary mobile metric turns into a question about how durable the company's entire business is.

Growth is no longer explosive

According to Sensor Tower, ChatGPT deletions rose 132% year over year in April 2026. A month earlier the figure was even harsher: up 413% year over year. For a product of ChatGPT's scale, deletions alone do not signal an immediate crisis, but they show that part of the audience has stopped seeing the service as the only option and has begun switching more actively between several AI assistants.

Growth in the active audience is slowing at the same time. In January, ChatGPT's monthly audience grew 168% year over year; in April it grew 78%. That is still strong momentum for an app of this size, and Sensor Tower estimates that ChatGPT keeps a substantially larger user base than its competitors. But the market looks not only at absolute size but at trajectory: if the category flagship loses pace, investors and partners start asking about the ceiling on growth.

Competitors are taking the momentum

The most uncomfortable signal for OpenAI is that the slowdown is not happening in a vacuum. Over recent months, ChatGPT downloads grew only 14% year over year, while Claude grew roughly 11-fold over the same period. That does not mean Claude has caught up with the leader in scale, but it does mean new users are far more willing to try alternatives than before. This already undercuts the familiar story of ChatGPT's unchallenged leadership in the consumer market.

An additional blow to the brand's image came at the end of February, when OpenAI signed a deal with the Pentagon. Sensor Tower separately noted the reaction of the US mobile audience, and it was telling: instead of ordinary user churn, the market saw a sharp spike in deletions, a dip in downloads and a rise in negative reviews. The cause matters beyond politics: for a mass-market app, such spikes mean that corporate news instantly affects user metrics.

• On February 28, ChatGPT deletions in the US jumped 295% day over day
• the average daily deletion rate before that was around 9%
• ChatGPT downloads first fell 13% after the news, then another 5%
• Claude's US downloads rose 37% and 51% on those same days
• the number of one-star ChatGPT reviews increased sharply

It is important that these are primarily mobile signals, not a full picture of OpenAI's entire user base. But such shifts are often the first to reveal a change in sentiment: people do not necessarily stop using ChatGPT altogether, but increasingly keep Claude, Gemini or other services close at hand and split their queries among them. For partners and investors, this signals that the AI chatbot market has become truly multi-platform.

Why this matters for the IPO

A dip in downloads rarely wrecks the story of a company of OpenAI's stature. The problem lies elsewhere: this data lands on top of reports that the company is already missing its own growth plans. According to The Wall Street Journal, as relayed by Reuters, OpenAI has fallen short of its internal targets in recent months on both revenue and new-user numbers. It was also separately reported that the company did not reach its internal goal of 1 billion weekly ChatGPT users by the end of 2025.

For a potential IPO, this set of metrics looks sensitive for several reasons. The market will look not at the loudness of the ChatGPT brand but at how reliably OpenAI can turn popularity into sustainable economics. Investors want a story not only about hype but about predictable demand, conversion into revenue and the ability to fund the company's infrastructure appetite. It is at such moments that analysts begin separating a strong product from a durable public company:

• does the service retain its audience as competitors get stronger
• is growth turning into revenue fast enough
• is the pace of the business enough to cover future compute contracts
• is the consumer segment becoming less predictable than it seemed a year ago

Reuters also reported that CFO Sarah Friar had voiced concern: if revenue does not grow fast enough, the company may struggle to pay for future compute commitments. OpenAI publicly rejected an overly dramatic reading of this story and insisted that leadership is united in its bet on continuing to expand computing capacity. But the very fact that such discussions are taking place shows that the company now cares not only about usage records but about the quality of growth.

What this means

For OpenAI this is not a catastrophe but a transition to a more mature phase of the market, where scale and brand recognition alone are no longer enough. For the whole industry the signal matters even more: the era of a single undisputed winner is ending, and the key metric is becoming not just install counts but retention, monetization and the ability to defend market share when users switch easily among several AI services and pick the best tool for each task.