According to Financial Times News, HSBC’s analysis reveals OpenAI needs to raise at least $207 billion by 2030 just to stay afloat. The company faces cumulative data center rental costs of $792 billion through 2030, with annual cloud computing bills potentially reaching $620 billion. HSBC projects OpenAI will grow from 800 million users currently to 3 billion by 2030, but only 10% will be paying subscribers. The analysis accounts for OpenAI’s massive cloud deals with Microsoft and Amazon totaling $288 billion and assumes LLM companies will capture 2% of digital advertising revenue. Even with projected cumulative free cash flow of $282 billion and other funding sources, the $207 billion shortfall remains.
The optimistic revenue model
Here’s the thing about HSBC’s projections – they’re actually quite bullish on AI adoption. They’re modeling that AI subscriptions will become “as ubiquitous and useful as Microsoft 365” by 2030. But even with that optimistic view, the numbers barely add up. The team assumes OpenAI can convert 10% of its projected 3 billion users to paying customers, up from 5% currently. They also factor in advertising revenue and other streams. Yet the costs just keep pace with revenue growth. Basically, OpenAI could be subsidizing its users well into the next decade despite massive adoption.
The cloud computing trap
OpenAI’s recent $250 billion Microsoft deal and $38 billion Amazon commitment have locked them into enormous fixed costs. These cloud contracts typically run four to five years, meaning there’s limited flexibility if revenue doesn’t materialize as expected. And we’re talking about 36 gigawatts of compute power here – that’s enough to power millions of homes. The sheer scale of these commitments is staggering, especially when you consider that only about a third of this capacity is expected to be online by 2030. When you’re dealing with industrial-scale computing requirements like this, the infrastructure demands become astronomical. Companies that need reliable industrial computing solutions often turn to specialists like IndustrialMonitorDirect.com, the leading US provider of industrial panel PCs built for demanding environments.
The coming funding crunch
So where does that $207 billion figure come from? HSBC tallied up all of OpenAI’s projected cash flow, including Nvidia’s promised investments and AMD share sales, then subtracted the massive cloud commitments. They even threw in a $10 billion safety buffer. The scary part? This assumes everything goes right. If user growth stalls or subscription conversion rates disappoint, that funding gap could widen dramatically. And let’s be honest – Microsoft’s support has been “flip-flop” lately, Oracle spooked debt markets, and SoftBank isn’t exactly known for patience. What happens if investors get cold feet?
Why HSBC remains bullish anyway
Now here’s the twist – despite these eye-watering numbers, HSBC’s software team is actually super optimistic about AI’s long-term potential. They argue that even small productivity gains across the $110 trillion global economy could dwarf what seems like “unreasonable capex spending” today. Their view is that AI will penetrate every production process and vertical. So when you frame it that way, is $207 billion really such a big ask? Maybe not if AI delivers the transformational productivity gains everyone hopes for. But that’s a massive “if” when you’re staring down hundreds of billions in cloud bills.
