The largest private funding round in history closed at $122 billion. Revenue hit $2 billion a month. The burn rate will hit $57 billion next year. And Anthropic is quietly winning the enterprise market OpenAI needs most.

OpenAI just closed $122 billion in committed capital at an $852 billion post-money valuation. That makes it the largest private funding round in history, more than doubling its own previous record from February. The round was anchored by Amazon ($50 billion), NVIDIA ($30 billion), and SoftBank ($30 billion), with continued support from Microsoft and participation from a16z, D.E. Shaw Ventures, MGX, TPG, and T. Rowe Price.

The revenue numbers behind the valuation are genuinely extraordinary. OpenAI is generating $2 billion in revenue per month. At the end of 2024, it was doing $1 billion per quarter. That is an 8x increase in roughly 15 months. ChatGPT has 900 million weekly active users, 50 million subscribers, and six times the web visits of the next largest AI app. The company claims it is "growing revenue four times faster than the companies who defined the Internet and mobile eras."

And it has never turned a profit.

The investor list reads like a who's who of technology infrastructure. It should. These are not passive financial investors. They are OpenAI's own suppliers and customers.

NVIDIA invested $30 billion. NVIDIA also sells OpenAI the chips that power its models. OpenAI is committing to 3 GW of dedicated inference capacity and 2 GW of training capacity on NVIDIA Vera Rubin systems. Jensen Huang told CNBC this "might be the last" pre-IPO investment. It might also be one of the best customer acquisition deals in history: invest billions in a company that spends billions buying your product.

Amazon committed $50 billion, though $35 billion of that is contingent on OpenAI's IPO proceeding or on the company reaching an AGI milestone. Alongside the investment, Amazon expanded its existing cloud agreement by $100 billion over eight years. AWS becomes the exclusive third-party distribution provider for OpenAI Frontier, the enterprise platform. William Blair analysts estimate the deal could generate roughly $17 billion per year for AWS, about 11 percent of expected 2026 AWS revenue. Amazon is not just investing in OpenAI. It is buying a customer.

Microsoft continues to support OpenAI while receiving a 20 percent revenue share under their partnership. At $2 billion per month, that share is worth $400 million monthly. Microsoft is also OpenAI's primary cloud provider. The relationship generates revenue in both directions.

Veteran tech commentator Om Malik published a piece on March 31 titled "OpenAI: The Fix Is In," arguing that the deal structure is designed to prop up the valuation ahead of an IPO. The Register flagged the circularity directly: investors supply the infrastructure OpenAI buys with the capital those same investors provide.

None of this means the revenue is fake. $2 billion a month from 900 million users and 9 million paying business customers is real demand. But the financial relationships between OpenAI and its largest investors create a feedback loop that makes the growth numbers harder to evaluate at face value. When your chip supplier, your cloud provider, and your revenue-sharing partner are all on your cap table, the line between investment and commercial arrangement gets blurry.

Here is where the story gets complicated for OpenAI. Consumer dominance is not in question. Enterprise is.

Enterprise revenue now accounts for more than 40 percent of OpenAI's total, up from 20 to 25 percent in 2025. The company says it is on track for enterprise to reach parity with consumer by the end of 2026. Nine million businesses pay for OpenAI products. The API processes 15 billion tokens per minute. Codex has 2 million weekly users, up 5x in three months.

But the market share data tells a different story. According to Ramp's spending data, OpenAI's share of enterprise AI spending fell from 50 percent to 27 percent over the past year. Anthropic's share climbed to 40 percent. Among companies buying AI tools for the first time, Anthropic captures 73 percent of spending. Epoch AI estimates that Anthropic could surpass OpenAI in annualised revenue by mid-2026.

That shift matters because enterprise revenue is stickier and higher-margin than consumer subscriptions. A company that integrates Claude into its compliance workflows does not switch to ChatGPT because of a price cut. The switching costs are real. Anthropic now has 500+ enterprise customers spending $1 million or more per year and has penetrated 70 percent of the Fortune 100.

OpenAI's $852 billion valuation is built primarily on consumer dominance: 900 million users, 6x the web traffic of competitors, the most recognisable AI brand in the world. The question is whether consumer dominance alone justifies that valuation, or whether the enterprise market, where the margins and stickiness live, is the one that ultimately determines who wins.

OpenAI's answer to the enterprise challenge is consolidation.

On March 19, the Wall Street Journal reported that OpenAI plans to merge ChatGPT, Codex, and its Atlas browser into a single desktop superapp. The internal memo, from Fidji Simo, CEO of Applications, frames it as reducing fragmentation across separate products and creating a platform where AI agents can coordinate across tools to complete complex tasks autonomously.

The strategic logic is clear. If 900 million weekly users interact with agents through a single app that handles code, browsing, data analysis, and task execution, OpenAI becomes the default operating layer for AI-powered work. That is a platform, not a product. And platforms command platform valuations.

The risk is equally clear. Every agent capability bundled into an app with 900 million users is an attack surface multiplied by that user base. As we mapped in our analysis of AI agent security evidence, every agent tested in red-team exercises was compromised at least once. Credit card exfiltration succeeded 10 out of 10 times. OpenAI itself has stated that prompt injection is "unlikely to ever be fully solved." Shipping autonomous agents to 900 million users while that statement stands is a bet that the commercial upside outweighs the security risk.

Strip away the growth narrative and the financial picture is sobering.

OpenAI projects $14 billion in losses for 2026. Cash burn is expected to hit $25 billion this year, rise to $57 billion in 2027, and peak at $47 billion in 2028. Cumulative cash burn through 2030 is projected at $665 billion. HSBC analysts estimate the company still needs to find $207 billion to meet its data centre commitments. One analyst told Tom's Hardware that OpenAI could run out of cash by mid-2027 without additional funding.

The IPO is not a strategic choice. It is a financial necessity. Amazon's $35 billion tranche is contingent on it happening. CFO Sarah Friar has been building the finance team for it, hiring Ajmere Dale from Block and Cynthia Gaylor from DocuSign. She told CNBC that ChatGPT must become a "productivity tool" before going public. The target is a valuation near $1 trillion.

For context, Anthropic is eyeing a $60 billion IPO as soon as Q4 2026. Its burn rate is forecast to drop to roughly 33 percent of revenue this year, then 9 percent by 2027. OpenAI's burn rate is heading in the opposite direction.

The $852 billion valuation prices in a future where OpenAI becomes the default infrastructure layer for AI. The consumer numbers support that thesis today. The enterprise numbers are wobbling. The financial trajectory depends on an IPO that has not happened yet, infrastructure deals with investors who are also suppliers, and a path to profitability that is still four years away.

The Bank of England has warned of growing risks of a global market correction due to possible overvaluation of leading AI firms. Wikipedia now has a dedicated "AI bubble" article. OpenAI has more than tripled its valuation in 18 months, from $157 billion in October 2024 to $852 billion today.

None of that means OpenAI is overvalued. $2 billion a month in revenue from a product that 900 million people use every week is not a mirage. But the gap between what the valuation assumes and what the financials show is the widest it has ever been. The next 18 months, the IPO, the superapp launch, the enterprise market share fight with Anthropic, will determine whether $852 billion was the price of conviction or the peak of a cycle.

If OpenAI's largest investors are also its largest suppliers, how should we read a valuation built on revenue that flows back to the companies funding it?