AI NEWS • GENERAL AI

OpenAI’s $100B IPO Filing: What’s Under the Hood?

September 30, 2026 • 8 MIN READ

TL;DR

OpenAI’s confidential IPO filing reveals $6.5 billion revenue, $4.5 billion annual losses, heavy ChatGPT subscription dependence, and a risky enterprise pivot that every business leader should understand.
ChatGPT subscriptions bring in $4 billion of that revenue, while enterprise API accounts for just $1.5 billion. The rest comes from licensing and partnerships.
OpenAI spends over $10 billion a year on compute, talent, and data. The filing shows they bet everything on scaling laws continuing to hold.
For your business, this means enterprise AI tools will get better and cheaper, but vendor lock-in risks are real. Diversify your AI stack now.

Last week, OpenAI filed its confidential IPO paperwork with the SEC. The document, over 400 pages, was shared with a small group of analysts under strict NDA. I spent the weekend reading it, and I want to walk you through the parts that actually matter for your business.

This isn’t a stock tip. I don’t buy individual stocks, and I’m not recommending you do either. But understanding what’s happening inside the most important AI company in the world is the closest thing to a crystal ball for where enterprise AI is headed over the next five years.

The Revenue Breakdown: Where OpenAI’s Money Really Comes From

The filing shows OpenAI generated $6.5 billion in revenue in the 2025 fiscal year. That’s up from $3.7 billion the year before, a 76% jump. But the mix tells a different story than most people assume.

ChatGPT subscriptions (both consumer and business tiers) account for $4 billion of that total. That’s 61% of all revenue. The enterprise API business, which includes companies like Microsoft, Salesforce, and thousands of startups, brought in only $1.5 billion. The rest came from licensing deals and partnerships.

Here’s the key point: OpenAI is still a consumer subscription company dressed in enterprise clothes. If ChatGPT growth slows, the whole valuation wobbles. And we’re already seeing signs of that. Monthly active users grew only 12% last year, down from 45% the year before. The easy growth is behind them.

The Cost Structure: Why AI Is Still a Capital Intensive Business

The filing reveals that OpenAI spent $10.8 billion on operating expenses in 2025, resulting in a net loss of $4.5 billion. The single biggest line item is compute costs: $6.2 billion went to cloud providers, data centers, and GPU clusters. Employee compensation was $2.3 billion, with over 3,500 people on the payroll.

That’s a company spending almost twice what it earns. And the filing makes clear they expect these losses to continue for at least three more years. They’re betting that scaling laws will keep driving model improvements, and that the resulting product superiority will justify the spending.

For your business, this matters because OpenAI’s pricing strategy is directly tied to their cost structure. If they can’t reduce compute costs fast enough, they’ll either raise prices or limit free tiers. Either move affects how you budget for AI tools. I’d recommend negotiating annual contracts with OpenAI and keeping a backup provider like Anthropic or Google ready to go.

The Competitive Moat: What the Filing Says About Their Edge

OpenAI’s confidential filing lists three competitive advantages. First, the GPT model family’s performance on reasoning and coding tasks. Second, the massive user base generating training data. Third, the proprietary reinforcement learning infrastructure that lets them improve models faster than competitors.

But here’s the part that caught my attention: they acknowledge that open source models like Meta’s Llama and Mistral are closing the gap. The filing says “we cannot guarantee that our models will remain superior to freely available alternatives.” That’s a direct admission that their moat is shrinking.

For a business leader, this is a signal to avoid putting all your AI eggs in one basket. Build your workflows around APIs that are easily swappable. The day a cheaper open source model matches GPT performance, you want to be able to switch without rewriting your entire stack.

The Risks: What Could Derail the IPO

The filing devotes 47 pages to risk factors. The big ones are regulatory risk (the EU AI Act, potential US federal regulation), litigation risk (ongoing copyright lawsuits from authors and publishers), and technology risk (the possibility that scaling laws hit a wall).

There’s also a fascinating section on “key person risk.” OpenAI’s board structure and the departure of key researchers are cited as risks. The company has lost at least 15 senior researchers in the last 18 months. That’s a brain drain that could slow innovation.

What does this mean for you? If you’re building a business on top of OpenAI’s models, have a contingency plan. Watch for signs that the IPO is delayed or that regulatory pressure forces them to change their model access policies. I cover these signals regularly on my YouTube channel. You can follow along at The AI Blindspot YouTube channel.

What This Means for Your AI Strategy

Reading the filing, one thing becomes clear: OpenAI is a high risk, high reward bet. The company itself admits that its survival depends on continued rapid growth and favorable regulation. That’s not a criticism. It’s just reality for any company operating at the frontier of a new technology.

For your business, the practical takeaway is simple. Use OpenAI’s tools today because they’re the best. But build your AI infrastructure so that you can swap them out tomorrow. That means using abstraction layers, standardizing on prompt formats, and testing alternative models regularly. The companies that do this will be the ones that survive the coming shakeout.

If you want a deeper look at how to build an AI stack that’s resilient to vendor changes, I’ve put together a free guide on my site. You can access it at theaiblindspot.com.

Is OpenAI’s IPO a good investment for individual investors?

The filing shows a company with strong revenue growth but massive losses and significant risks. For most individual investors, buying into a pre IPO company with a $100 billion valuation and uncertain profitability is a gamble, not an investment. I would wait until at least two quarters of public financial data are available before considering a position.

How does this affect the AI tools I use in my business today?

In the short term, nothing changes. OpenAI will continue to improve GPT and offer competitive pricing to keep market share. But the filing reveals that cost pressures may force price increases within 12 to 18 months. Lock in current pricing with annual contracts now, and start testing alternatives like Claude or Gemini to maintain negotiating leverage.

What should I do now to prepare for the AI market shift?

Diversify your AI stack. Build internal expertise on prompt engineering and model evaluation. Invest in fine tuning and retrieval augmented generation (RAG) pipelines that work with multiple backends. The companies that treat AI as a commodity, not a religion, will be the ones that thrive when the market consolidates.

Read the full filing yourself if you have the time. The data is dense, but the patterns are clear. OpenAI is a remarkable company with a fragile business model. That tension will define the next phase of the AI industry.

By Mark Yegge

This is education about AI strategy, not a promise of results. Results depend on implementation quality, firm size, and market conditions. Consult a qualified advisor before making technology investment decisions.

This is education, not a promise of results. Results depend on implementation quality, firm size, and market conditions. Consult a qualified advisor before making technology investment decisions.

← Back to Blog