OpenAI is the most talked-about company in AI, and its revenue numbers are watched like a stock ticker. So it turned heads this month when a new report revealed that OpenAI is bringing in billions of dollars less than many people believed. The twist is that the gap has almost nothing to do with how the business is actually performing.
The $50 billion reality
According to the Financial Times, OpenAI told investors that its annualized revenue was near $50 billion at the end of September. That is a huge number, but it is roughly $20 billion below a figure of around $70 billion that had been circulating just weeks earlier. CNBC confirmed the lower figure the same day.
Annualized revenue is a run rate: it takes recent monthly sales and projects them across a full year. So this is not a report of OpenAI losing money or shrinking. It is a correction to how much money the company is actually on track to make, and the real number is tens of billions lower than the hype suggested.
Why the numbers did not match
Here is the part that matters most: the difference is accounting, not demand. OpenAI counts its revenue net, meaning it leaves out sales made through its cloud partners. The higher $70 billion figure appears to have come from OpenAI's own investors adjusting the number upward to include those partner sales, so it could be compared directly against rival Anthropic, which reports its revenue that way.
In other words, two companies were being measured with two different rulers, and someone tried to make them match. When you strip that out, OpenAI's honest run rate is about $50 billion. And the underlying business is still growing quickly: OpenAI has told investors its revenue has grown more than 70% in the period.
What it means for you
This story is a useful reminder to read AI headlines carefully. The AI industry runs on enormous, attention-grabbing numbers, and those numbers are not always measured the same way. A $20 billion swing that comes purely from an accounting choice shows how easily hype can inflate the figures everyone repeats.
For business leaders making decisions based on AI market claims, the practical advice is to ask what is actually being counted before you treat a number as fact. Revenue, growth rates, and user counts can all be framed to look bigger. The companies worth betting on are the ones whose results hold up when you look at how the numbers are built, not just how big they sound.
OpenAI is still a juggernaut growing at a pace most companies can only dream of. But the $50 billion correction is a healthy dose of reality in an industry that often forgets the difference between a real number and an impressive one.
