
The report cites a recent analysis of OpenAI’s finances from British multinational financial services giant HSBC, which took into account the AI startup’s planned spending on infrastructure, compute and energy costs, as well as its projected revenues to offset all those costs.
The bank estimates that OpenAI will bring in a bill of $620 billion a year on data center costs, with the caveat that it has signed contracts for more computing power than is actually available at the moment. It then made a projection for the company’s customer reach, which is currently said to be 800 million according to OpenAI’s numbers and will reach three billion by 2030 under HSBC’s model. The bank generously estimates that OpenAI will convert 10% of that reach into paying customers, double its current rate of 5%. Those estimates are more generous than OpenAI’s reported internal projections, which pegged the company at $2.6 billion and converting 8.5% of them into paying customers by the end of the decade. HSBC has also reduced some of OpenAI’s advertising revenue under the assumption that AI companies will account for about 2% of the total digital advertising market in the coming years.
With all this, HSBC estimates that OpenAI’s annual revenue will reach approximately $215 billion by 2030. This, once again, tops OpenAI’s own projections, which reportedly put it at around $200 billion annually by the end of the decade. Both models are asking for basically unprecedented growth, but let’s get on with it. Taking into account OpenAI’s current cash flow and its projected optimistic growth projections, the company will still be left with a funding deficit of $207 billion. According to HSBC, the company will need to raise this amount to continue operating at a loss.
OpenAI has options for closing that funding gap, though none of them are as attractive. The company may step back from some of its data center commitments to reduce its expenses, though that may not provide much comfort to investors who are expecting something close to infinite growth. It could also beat the generous revenue projections made by HSBC, which seems unlikely and not really something it can build on. If it were easy to generate revenue, the company would be doing it.
Then there’s the other option that OpenAI executives quickly floated before backing down: get a government bailout. Contingency plans are generally not a bad idea, but it probably doesn’t inspire much confidence that you’re planning so hard on the possibility of failing that you could take the entire economy down with you.
<a href