Greg Brockman, co-founder and president of OpenAI, made a statement at the WSJ Tech Conference in Napa, California, last Wednesday that reverberated well beyond the stage: the company now generates the majority of its revenue from enterprise sales, and he believes it is ready to go public in 2027. The answer to a direct question was concise — "I think so" — but behind those three words lies the most radical transformation OpenAI has faced since its founding as a non-profit research laboratory in San Francisco.
The story Brockman told was not one of a sudden pivot, but of a gradual and painful recalibration. The first two quarters of 2026 were, in the executive's own words, "pretty rough." OpenAI was not even "in the conversation" when it came to agentic coding — the territory that Anthropic had already conquered with Claude Code. It was a rare and revealing admission that, despite the ChatGPT halo, the company was falling behind in one of the most promising segments of the corporate AI market.
What changed in the third quarter was both dramatic and quiet. Annualized revenue growth jumped 70% quarter over quarter. But the more revealing figure is not the number itself — it is the composition of that revenue. The enterprise segment, which previously accounted for about 30% of the total, now represents the majority. "It's not that consumer revenue has shrunk," Brockman emphasized. "It's been growing. But the enterprise segment has really taken off."
This semantic detail is crucial. To an outside observer, OpenAI remains the ChatGPT company — more than 900 million weekly active users, more than 50 million paid subscribers. What actually occurred was a structural shift in revenue, not an abandonment of the consumer market. Enterprises like JPMorgan Chase, which Brockman cited as a recent partnership example, pay significantly more than individual consumers, and this leverage effect transformed the company's financial equation.
The Codex, the programming agent that OpenAI had been promoting quietly, is the centrepiece of this strategy. With over 2 million weekly users after a fivefold expansion in three months and activity growth exceeding 70% month over month, Codex represents what Brockman described as "head-down execution." While the industry focused on Anthropic's Claude Code, OpenAI was building its own agentic productivity tool, without noise, until the numbers spoke for themselves.
The race to an IPO is, in part, a response to competitive pressure from Anthropic, which according to Reuters is likely to list before the end of 2026. OpenAI already filed a confidential S-1 with the SEC in June 2026 — the first formal step toward a public offering — but concluded that 2026 was too early. The timeline was reset to 2027, but without a concrete date, a defined exchange, or an offering price.
It is worth noting that the numbers behind this transformation are not modest. OpenAI generates approximately 2 billion dollars per month, according to data published earlier in 2026. With enterprise revenue now dominant and accelerating growth, the company finds itself in a position that few outside observers had predicted. In March 2026, OpenAI itself announced that enterprise revenue was on track to reach parity with consumer revenue by the end of that year. What happened after that announcement was unexpected: enterprise revenue not only caught up with consumer revenue, but surpassed it.
The question that remains is how much of this transformation reflects genuine market demand versus the maturity narrative the company has constructed for its investors and the public markets. If OpenAI truly needed to compete with Anthropic to win enterprises — suggesting that the conquest was not automatic but the result of deliberate sales efforts and partnerships — then product superiority may still be more marketing myth than consolidated reality.
Sources: TechCrunch, Unite.AI, CryptoBriefing
✓ Independent sources cross-checked and verified before publishing
