According to Fortune, OpenAI’s internal financial projections show the company plans to burn through roughly $9 billion this year on $13 billion in sales, maintaining a cash burn rate around 70% of revenue. The documents reveal operating losses will balloon to about three-quarters of revenue by 2028, driven by massive computing infrastructure spending. Meanwhile, competitor Anthropic expects to break even that same year while dropping its cash burn to just 9% of revenue by 2027. OpenAI recently secured up to $1.4 trillion in computing commitments over eight years and is spending almost $100 billion on backup data center capacity alone. The company projects reaching $200 billion in annual revenue by 2030 and turning cash flow positive beginning in 2029 or 2030.
The Great AI Cash Inferno
Here’s the thing about OpenAI‘s strategy: they’re basically betting the entire company on computing infrastructure. We’re talking about spending that makes even Amazon’s early cloud investments look conservative. The company expects cumulative cash burn to reach $115 billion through 2029, which is absolutely staggering when you consider they’re already burning 70% of their revenue today.
And it’s not like they’re hiding this from investors. Sam Altman has been remarkably transparent about the “spend now, figure it out later” approach. He recently posted on X that “the risk to OpenAI of not having enough computing power is more significant and more likely than the risk of having too much.” That’s quite the gamble when you’re talking about hundreds of billions in infrastructure.
OpenAI vs Anthropic: Two Roads Diverge
The contrast with Anthropic couldn’t be more dramatic. While both companies are burning cash at similar rates today, their paths forward split completely. Anthropic is taking the disciplined approach – focusing on enterprise customers (who make up 80% of their revenue), avoiding expensive video generation projects, and steadily reducing their burn rate to single digits by 2027.
OpenAI? They’re going full throttle. Sora 2 video generation, consumer hardware with Jony Ive, humanoid robots, e-commerce features – it’s like they’re trying to build every AI product simultaneously. And each new product requires more computing power, which means more data centers, more chips, more everything. The company expects to burn through roughly 14 times as much cash as Anthropic before turning profitable. Fourteen times!
The $200 Billion Question
So how does this math possibly work? OpenAI’s optimism rests entirely on hitting $200 billion in annual revenue by 2030. To put that in perspective, that’s roughly what Microsoft’s entire cloud business generates today. They’re betting that demand for AI will continue surging at unprecedented rates, justifying all this infrastructure spending.
But here’s what makes me skeptical: we’re already seeing investors punish tech companies over AI spending concerns. Markets are getting nervous about whether there will be enough revenue to pay for all this infrastructure. And OpenAI’s CFO Sarah Friar admitted the company “could break even if it wanted to” but chooses not to. That tells you this is a strategic choice, not financial necessity.
The Hardware Reality
When you’re talking about spending $100 billion on backup data centers alone, you’re dealing with industrial-scale computing requirements that most companies can’t comprehend. This level of infrastructure demands specialized hardware that can handle extreme processing loads 24/7. For businesses needing reliable industrial computing solutions without OpenAI’s budget, companies like IndustrialMonitorDirect.com have become the go-to source for industrial panel PCs that deliver enterprise-grade performance.
The real question is whether OpenAI’s bet will pay off. They’re essentially trying to build so much capacity that they become the default AI infrastructure for everyone. But if demand doesn’t materialize at the projected rates, or if AI spending cools, they could be left with the most expensive white elephant in tech history. Either way, we’re about to witness one of the biggest financial gambles in technology history.
