OpenAI’s CFO Sets a 2027 IPO Target as Enterprise Takes the Lead

OpenAI has moved its potential IPO from market speculation toward a timetable. CFO Sarah Friar told employees the company will be public in 2027, or earlier if the business keeps improving—a target that puts its enterprise growth, revenue quality and cost structure under a far brighter investor spotlight.

The formal process had already begun. OpenAI confidentially filed an IPO prospectus with the U.S. Securities and Exchange Commission in June, though it has not announced a public debut date. Friar told employees that a March financing round raised $122 billion and gave the company flexibility, framing an IPO as another financing milestone rather than an endpoint.

The distinction matters. A confidential filing allows a company to prepare for public markets without committing to a launch. Friar’s condition—an earlier debut only if the business continues to improve—makes operating results part of the timetable. OpenAI is currently described as carrying an $852 billion valuation, raising the standard for the financial disclosure investors will eventually receive.

At a recent shareholder meeting, Friar said enterprise revenue had surpassed consumer revenue. That reported crossover is consequential for the IPO narrative: OpenAI can point to business adoption rather than relying solely on its consumer products. It also arrived ahead of the company’s reported expectation of reaching revenue parity between the two segments by the end of 2026.

35% quarter to dateRevenue run-rate growth

OpenAI told employees its revenue run rate was up 35% quarter to date.

50% quarter to dateEnterprise run-rate growth

OpenAI told employees its enterprise revenue run rate was up 50% quarter to date.

20 million weekly active usersAI coding and work users

OpenAI reported 20 million weekly active users for its AI coding and work product.

OpenAI’s reported $40 billion annualized revenue run rate, including 20% month-over-month growth in July, is the headline figure in that case. The company has also said its AI coding and work product reached 20 million weekly active users. Advertising in ChatGPT was reported to be nearing a $1 billion annualized run rate, adding a smaller but potentially distinct revenue line.

Annualized revenue is a projection from a shorter period, not a completed year of sales. One account says OpenAI has calculated the measure by multiplying revenue from its most recent four weeks by 12. That can be a useful directional measure of momentum, but it does not by itself establish the stability of demand across a full year.

The more conventional quarterly figure points to a slower, more qualified picture. People familiar with the matter told The Wall Street Journal that OpenAI recorded $6.7 billion in revenue in the June quarter, up from $5.7 billion in the prior quarter—an 18% increase. The Journal also reported that some shareholders were disappointed by the company’s progress relative to Anthropic.

The disclosures investors will press for

• Whether the enterprise crossover translates into durable contracted revenue rather than a short-term acceleration in usage.
• How revenue growth compares with the costs of serving customers and training models. One reported Q1 figure put revenue at $5.7 billion and cost of revenue plus training at $12.1 billion.
• Whether recent growth after the public launch of the GPT-5.6 model family persists beyond the launch period. OpenAI reportedly told investors that growth accelerated after that release.

Anthropic may reach public markets before OpenAI. Friar told employees that Anthropic could disclose its own confidential filing and become public as early as September, while emphasizing that OpenAI would follow its own process. That possibility means OpenAI may not get to define the valuation framework for frontier AI companies on its own.

OpenAI is also approaching the possible listing amid senior-level change. The Financial Times reportedly counted nearly half a dozen reorganizations during 2026, including executive departures and role changes. It also reported that employees had been cashing out through a tender offer of nearly $7 billion at the $852 billion valuation. Those steps can give staff liquidity before an IPO, but they also ensure that organization and governance will be part of the public-market examination.

Friar has supplied a window, not a final date. Between now and 2027, the essential test is whether OpenAI can turn its enterprise-led growth claims into financial results that withstand public scrutiny—while demonstrating how quickly revenue can grow relative to the costs behind it. The confidential filing gives OpenAI room to choose its moment; the reported numbers will determine how much room investors allow.