OpenAI: Enterprise Revenue Surpasses Consumer for the First Time with $40 Billion ARR
CFO Sarah Friar confirms to investors the inversion: enterprise grew 32% in one month and surpassed consumer four months earlier than expected, with an IPO above $1 trillion on the horizon.
Sarah Friar, CFO of OpenAI, announced to investors on August 14 that the revenue from the enterprise segment has for the first time surpassed that of consumers in the company’s history, four months ahead of what Friar had projected at the beginning of 2026. The annual revenue rate reached $40 billion, and revenue from corporate customers grew by 32% in just one month, according to CNBC. The company entered the year with 60% of revenue coming from consumers and 40% from enterprise; the inversion happened faster than the internal growth model had estimated.
"We started the year at 60-40, but enterprise accelerated much more quickly than expected, and the lines have already crossed," Friar told investors, according to CNBC’s report. Advertising within ChatGPT, initiated in February 2026, is approaching $1 billion in ARR, a third revenue vector that did not exist 18 months ago.
The Inversion and What It Changes
Enterprise contracts create high switching costs, predictable renewal cycles, and account expansion potential that the consumer base does not offer. The volatility of monthly users, which has dominated media coverage over the past two years, loses relevance as a health indicator of the business. What now matters is the Net Revenue Retention from contracts via Azure OpenAI Service and ChatGPT Enterprise.
Microsoft, with an accumulated investment of over $13 billion, distributes OpenAI models via Azure to corporate clients in over 50 countries. The shift to enterprise at OpenAI is built partly on this distribution network, creating mutual dependence that extends beyond a conventional investment agreement. For customers using Copilot for Microsoft 365, integration with OpenAI models is often invisible, making the switching cost higher than the customer realizes when signing the initial contract.
What the IPO Changes for AI Buyers
OpenAI confidentially filed with the SEC in June 2026, with Goldman Sachs and Morgan Stanley as lead banks and a target valuation above $1 trillion. The disclosure of financial data by Friar to a group of prospective investors is part of the preparation for a roadshow: the company is building the argument that the enterprise base is stable and growing, which reduces the perceived risk of betting on a private company whose fundamentals were opaque.
2027 renewals will occur with a public counterpart, under pressure for quarterly revenue growth, with less willingness for the entry discounts that characterized the first enterprise agreements of 2024 and 2025. Companies that negotiated three-year contracts at 2024 prices are now in the most favorable position they will have.
Anthropic signals an IPO in the same fall, with preliminary revenue of $11.5 billion in Q2 2026. Two AI laboratory IPOs in the same period create public comparison pressure on multiples: the market will price both in real time, and revenue retention margins below expectations will immediately reflect in stock prices.
Japan at the Forefront, Europe on Alert
Japan is the most advanced enterprise adoption market outside the U.S. MUFG, Nomura, and Sony are among the institutions that have closed institutional licenses for ChatGPT Enterprise since early 2025, driven by the corporate digitalization policy of METI. The confirmation that enterprise is OpenAI's largest revenue line is read in Tokyo as validation of an early bet: Japanese banks and corporations that invested in training and integration before 2026 have a relative advantage over Europeans, who have progressed more slowly due to GDPR restrictions and uncertainty regarding the AI Act.
In Europe, the AI Act entered the enforcement phase for general-use models on August 2, 2026, requiring OpenAI, as a provider of systemic impact models, to maintain detailed technical documentation and be available for auditing by the EU AI Office. An IPO above $1 trillion increases the political visibility of the company, and the pressure to demonstrate that the AI Act is effective grows in direct proportion to the size of the target.
For consultancies advising European clients on OpenAI contracts, this implies concrete action: including regulatory compliance clauses in the 2026 and 2027 renewals, with explicit provisions for model auditing and data localization. The cost of including these clauses now, while the company is still private, is lower than the cost of renegotiating after the IPO, when bargaining power shifts.