OpenAI’s chief financial officer told investors and staff this week that the company plans to complete an initial public offering in 2027, with the date potentially moving earlier if revenue growth accelerates, according to people who attended the meeting. The statement is the firmest timeline the company has given for going public, ending years of speculation about when the most valuable startup in technology would test the public markets.
The CFO, Sarah Friar, said OpenAI’s second-quarter revenue reached $6.7 billion, up 18 percent from the first quarter, a pace the company believes supports a public listing within the next two years. The figure covers the company’s core businesses, including ChatGPT subscriptions and the sale of computing capacity and model access to developers and enterprises. OpenAI has been growing quickly, though it remains loss-making once the cost of training and running its models is counted.
The comparison that hung over the meeting was with Anthropic, OpenAI’s closest rival. Anthropic has reported preliminary second-quarter revenue of $11.5 billion, an increase of more than 140 percent from the prior quarter, according to figures circulated to its own investors. The gap is striking: the challenger is now booking more revenue than the company that invented the product category, and OpenAI’s team is under pressure to explain why its growth rate trails the competition.
The IPO plans come with a demanding price tag attached. OpenAI has been valued privately at $852 billion, and investors in the public markets will expect the company to justify that number with detailed financial disclosures it has so far refused to provide. The company filed its IPO registration confidentially with the Securities and Exchange Commission in June, a step that allows it to prepare quietly, but it has not said when it will begin the formal process or how large the offering will be.
The pressure is compounded by the state of the company’s leadership. OpenAI has seen a wave of senior departures in recent months, with executives across research, product, and operations leaving for other ventures or for competitors. Chief Executive Sam Altman has publicly acknowledged the churn and has been working to stabilize the management ranks, and people close to the company say the IPO calendar is now a factor in those efforts: a clean, senior team is considered essential to a successful listing, and the departures have raised questions about who will be running the company when it goes public.
The financial picture OpenAI presented to investors is more complex than the top-line number suggests. The company’s revenue is growing, but its costs grow with it, because every new customer using its models consumes computing power that must be bought in advance. OpenAI has secured large commitments from cloud providers, and it has struck deals to diversify its supply, but the economics of selling AI services remain unusual: the company’s gross margin depends on how efficiently it can run models at scale, a problem that has no settled answer.
The confidential filing, first reported in June, put OpenAI on a path that rivals are also following. Anthropic has filed its own confidential IPO paperwork and has begun meeting with potential investors, and the two companies are now racing to be the first large AI firm to list. A successful OpenAI offering would be one of the largest technology IPOs in American history, and it would test whether the public markets can absorb a company whose product is changing as fast as its revenue.
The executive departures have given the company’s critics ammunition. Skeptics argue that a firm losing senior talent at this rate is not ready for the scrutiny of quarterly earnings, where every departure becomes a story and every miss becomes a headline. Supporters counter that OpenAI’s growth justifies the pace, and that the company’s bench of managers is deeper than it appears. Both sides agree on one thing: the next two years will determine whether OpenAI can make the transition from private darling to public company without stumbling.
Friar’s comments also addressed the shape of the offering. The company is expected to sell a small fraction of its shares, with most of the proceeds going to existing shareholders rather than to the company itself, a structure that rewards early investors and employees. OpenAI has a complex ownership arrangement with Microsoft, which holds a large stake, and the IPO will require resolving how those shares are treated. People familiar with the process say the two companies have been in discussions about the mechanics for months.
The timing question is fundamentally a question about growth. If OpenAI’s revenue continues to climb at current rates, the company could accelerate the timeline and go public sooner than 2027, Friar said, and it would likely do so to capture a favorable market. If growth stalls, or if the broader market for technology stocks sours, the company can delay, filing extensions that allow it to remain private while keeping the process alive. The flexibility cuts both ways, and investors will parse every quarter’s numbers for clues about which path the company is on.
For the AI industry, the OpenAI IPO would be a defining event. The company’s valuation, its financial disclosures, and the reception its shares receive would set the terms by which every other AI startup is judged. Anthropic’s own offering, if it comes first, will provide the first real test of how public investors price a business built on models that improve monthly. OpenAI’s executives say they are prepared for that scrutiny. The company’s calendar now says they have until 2027 to prove it.
