OpenAI just raised $122 billion. The investors know it won't be profitable until 2029.

On March 31, 2026, OpenAI closed a $122 billion funding round at an $852 billion valuation. The company also projects $14 billion in losses this year. Everyone is reporting the first number. Fewer are sitting with both at the same time.

Who wrote the checks

Amazon anchored the round with $50 billion. Nvidia committed $30 billion. SoftBank put in another $30 billion and co-led alongside Andreessen Horowitz, D.E. Shaw Ventures, MGX, TPG, and T. Rowe Price Associates. Microsoft continued its long-standing participation. Reported by TechCrunch, roughly $3 billion came through individual investors via bank channels — the first time OpenAI opened the round to retail money before an IPO.

The full participant list runs through much of institutional finance: BlackRock, Blackstone, Sequoia, Fidelity, Temasek, Thrive Capital, Coatue, and others. The credit facility, also expanded to $4.7 billion and currently undrawn, is backstopped by JPMorgan Chase, Goldman Sachs, Citi, Morgan Stanley, Wells Fargo, and six other global banks.

This isn't a tech round anymore. It's a financial system event.

What OpenAI actually looks like right now

The consumer product numbers are real. ChatGPT has 900 million weekly active users and over 50 million paying subscribers. OpenAI's own announcement states it generates $2 billion in monthly revenue, growing four times faster than Alphabet and Meta did at the same stage. Search usage inside ChatGPT has nearly tripled in a year. The ads pilot crossed $100 million in ARR in under six weeks.

Enterprise is accelerating too. It now accounts for more than 40% of total revenue, with the company projecting parity between consumer and enterprise by end of 2026. Codex serves over 2 million weekly users, up 5x in three months, growing more than 70% month over month. OpenAI's APIs are processing more than 15 billion tokens per minute.

"Our consumer scale becomes the front door for enterprise usage, as familiarity in daily life drives adoption at work." — OpenAI

That quote is doing real analytical work. The consumer product isn't just a consumer product. It's a sales funnel for the enterprise contracts that will eventually justify the valuation.

Where the $122 billion goes

Chips, data centers, and talent, in that order. OpenAI's infrastructure strategy now runs across Microsoft, Oracle, AWS, CoreWeave, and Google Cloud for compute; Nvidia, AMD, AWS Trainium, and its own custom chip in partnership with Broadcom for silicon; and Oracle, SoftBank Energy, and SoftBank Group for physical data centers.

To free up compute for what comes next, OpenAI reportedly killed Sora, its video generation model. The company also launched GPT-5.4 and announced a ChatGPT super app merging chat, Codex, search, and agentic capabilities into a single product.

The framing in the announcement is worth reading carefully. OpenAI describes consumer adoption, enterprise deployment, developer usage, and compute as "a reinforcing flywheel." That's not a product strategy. That's an infrastructure moat argument — the case that no one can catch up because the lead compounds.

The fine print on Amazon's $50 billion

Here's the number buried under the headline: $35 billion of Amazon's $50 billion commitment is contingent on OpenAI completing an IPO or reaching artificial general intelligence. If neither condition is met, Amazon keeps $35 billion in its pocket.

The largest anchor investor in the largest private funding round in history has a conditional clause on 70% of its commitment. That's not a minor footnote. It means the actual secured capital is considerably smaller than $122 billion, and it ties a significant portion of the round's legitimacy to events that are not guaranteed.

OpenAI will join ARK Invest's ARKK, ARKW, and ARKF funds ahead of any IPO, giving retail investors exposure before shares reach a public exchange. That's a meaningful structural shift: public market pricing dynamics on a private company.

Why it matters

The losses aren't a sign the model is broken. They're the cost of running at the frontier. Compute costs are growing faster than revenue, which is why this round is runway, not profit-taking. Profitability isn't projected until 2029. Every dollar raised between now and then is buying time and scale.

At $852 billion, OpenAI is the second most valuable private company in history, behind only the SpaceX-xAI merger. The comparison OpenAI makes to Alphabet and Meta is carefully worded: revenue trajectory at the same stage, not valuation-to-earnings. The phrasing matters because Alphabet and Meta became two of the most profitable companies on earth. OpenAI is betting the same path is available to them. Whether that's true depends entirely on whether the flywheel holds.

The harder question isn't whether OpenAI can reach profitability. It's what the concentration of this much capital in a single private AI company does to everything else in the ecosystem. Open-source projects, smaller AI labs, and government regulators now operate inside the gravitational field of an entity that can outspend most national AI budgets without drawing on its credit line.

When one company controls that much compute, market share, and investor alignment simultaneously, the question of who sets the norms for AI development stops being philosophical.

It becomes structural. And structures are very hard to renegotiate once they're load-bearing.

If the capital competition in AI has already been decided, is the race for governance rules the only one still open?

Originally published as an Instagram carousel on @recul.ai.