Gemini — and in more recent months, Claude — have had surges in popularity, and this is beginning to show in OpenAI’s numbers.
According to a new report by the Wall Street Journal, OpenAI has missed its own internal targets for both new users and revenue — a stumble that has raised concern within the company as it barrels toward a highly anticipated IPO. The shortfalls are not trivial: OpenAI failed to hit an internal goal of reaching one billion weekly active users for ChatGPT by the end of 2025, and missed multiple monthly revenue targets earlier this year.
The CFO Is Worried
Chief Financial Officer Sarah Friar has reportedly told other senior leaders that she is concerned the company might not be able to honor future computing contracts if revenue does not grow fast enough. Friar, alongside other executives, is now pushing for tighter cost controls and greater financial discipline — a notable shift in tone for a company that has been in aggressive expansion mode for years.
OpenAI’s board has also begun scrutinizing its data center deals, with some directors questioning CEO Sam Altman’s push to secure even more computing infrastructure at a time when revenue growth is under pressure.
The Competitive Picture
The timing is uncomfortable. OpenAI’s market share erosion has been building for over a year. In 2024, Anthropic’s Claude nearly doubled its share of the AI market, climbing from 12% to 24%, while ChatGPT’s dominance gradually thinned. The WSJ report specifically notes that OpenAI has lost ground to Anthropic in coding and enterprise markets — two segments that drive the kind of high-value, sticky revenue that matters most ahead of an IPO.
Claude’s momentum has only accelerated in 2026. Claude Code’s annualized revenue crossed $2.5 billion in February, more than doubling since the start of the year. Anthropic claimed 73% of spending among new AI tool buyers — a reversal from OpenAI’s 90% consumer market share in 2024.
Google’s Gemini has also made inroads, particularly in the consumer vertical. As per SimilarWeb data, Gemini now gets over 25% of traffic, compared to just over 50% for OpenAI. The AI landscape has gone from a one-horse race to a genuinely competitive market in under two years.
Spending Problem
OpenAI has committed to spending roughly $600 billion on data centers over the coming years — an extraordinary capital commitment that made sense when the company’s revenue trajectory looked near-vertical. That calculus is now more complicated. The company reported a loss of $5 billion in 2024 on $3.7 billion in revenue, and analysts project losses widening to $14 billion by 2026.
The irony is that OpenAI remains, by most measures, the most recognized AI brand in the world. ChatGPT has over 800 million weekly active users and the company is valued at over $800 billion. But valuation and cash flow are different things, and it is the latter that pays for NVIDIA contracts.
IPO Pressure
The misses land at the worst possible moment. An IPO is expected before the end of 2026, and public market investors will scrutinize unit economics with a rigor that private backers have not. Friar and the board are right to demand more discipline. The question is whether tighter controls and an IPO timeline can coexist with the kind of massive compute bets that Altman believes are necessary to stay ahead. If revenue doesn’t catch up to the ambition, something will have to give — and in that scenario, it is unlikely to be the ambition.