OpenAI, the San Francisco artificial intelligence company that was founded in 2015 as a nonprofit research laboratory dedicated to ensuring that artificial general intelligence benefits all of humanity, has surpassed twenty-five billion dollars in annualized revenue — a figure that the company disclosed in private communications with investors and that was confirmed independently by three individuals with direct knowledge of the company’s finances. The milestone places OpenAI among the fastest-growing technology enterprises in recorded commercial history, surpassing the revenue trajectory of Google, Facebook, and Amazon at equivalent stages of their development by a margin so substantial that the comparison strains the frameworks that technology analysts typically employ to contextualize growth.

The revenue is generated primarily through ChatGPT subscriptions, API access fees charged to the hundreds of thousands of developers and enterprises that build applications on OpenAI’s models, and a growing portfolio of enterprise contracts that include custom model deployment, fine-tuning services, and dedicated compute allocations. The consumer subscription business, anchored by ChatGPT’s approximately three hundred million monthly active users, contributes the largest single share, but the enterprise segment is growing at a faster rate and commands substantially higher per-customer revenue — a pattern that investors have identified as evidence that OpenAI’s commercial trajectory is evolving from a consumer phenomenon into an enterprise infrastructure provider, a transition that historically produces the most durable and valuable technology businesses.

Anthropic, the company founded in 2021 by former OpenAI researchers Dario and Daniela Amodei, is approaching nineteen billion dollars in annualized revenue, a figure that has not been publicly confirmed by the company but that has been cited in multiple reports based on investor communications. The combined annualized revenue of the two leading AI companies — approximately forty-four billion dollars — represents a market that did not exist at any meaningful commercial scale three years ago. The velocity of this revenue creation has no precedent in the technology industry, and the competitive dynamics between OpenAI and Anthropic have produced a pace of product improvement, price reduction, and capability expansion that has benefited customers while consuming capital at a rate that makes profitability, for both companies, a horizon rather than a destination.

The possibility of an OpenAI initial public offering, which the company’s leadership has discussed with investment banks and which could occur as early as late 2026, introduces a set of questions that the technology industry and the public markets have not previously confronted. OpenAI’s corporate structure — a capped-profit subsidiary controlled by a nonprofit board whose stated mission is the safe development of artificial general intelligence — is unlike any entity that has previously listed on a major stock exchange. The tension between the fiduciary obligations that a public listing imposes (the legal duty to maximize shareholder value) and the mission obligations that the nonprofit structure was designed to preserve (the commitment to develop AGI safely and for the benefit of humanity) is not theoretical; it is structural, and no amount of legal engineering can fully resolve the competing claims of shareholders who seek returns and a charter that subordinates returns to safety.

The company has been restructuring its corporate governance in anticipation of a public listing, a process that has generated controversy both internally and in the broader AI policy community. The original nonprofit board has been reconstituted, Sam Altman has received equity in the company for the first time since its founding, and the capped-profit structure — which was designed to limit investor returns to one hundred times their investment — has been the subject of negotiations whose outcome will determine whether OpenAI enters the public markets as a conventional corporation or as something genuinely novel in the history of American capitalism.

The revenue figures, the IPO preparations, and the corporate restructuring are elements of a single narrative: the transformation of artificial intelligence from a research domain funded by philanthropic and government capital into a commercial industry of staggering scale, governed by the incentive structures and accountability mechanisms of the capital markets. This transformation was inevitable from the moment AI models demonstrated capabilities that enterprises would pay for, but its speed has outpaced the development of the regulatory, ethical, and governance frameworks that the technology’s implications demand. An OpenAI that is publicly traded is an OpenAI that is subject to quarterly earnings calls, analyst expectations, competitive pressure to ship capabilities faster than safety review processes might counsel, and the relentless logic of a market that rewards growth and punishes caution.

The competitive landscape in which these revenue figures exist is itself a subject of consequence. OpenAI and Anthropic together command the majority of the frontier AI market, but they operate under fundamentally different theories of the relationship between commercial success and safety. OpenAI has pursued maximum distribution — putting its models in the hands of the largest possible number of users and developers, on the theory that broad deployment generates the data and feedback necessary to improve both capability and safety. Anthropic has pursued what it describes as responsible scaling — expanding access deliberately, publishing safety research openly, and accepting slower commercial growth as the price of maintaining the institutional capacity to manage the risks that its own technology creates. The public markets, when they render their verdict on OpenAI’s IPO, will be rendering a judgment not only on the company’s commercial prospects but on which of these theories the capital markets consider more valuable.

Twenty-five billion dollars in annualized revenue, generated by a company that is nine years old and that began as a nonprofit, in a market that did not exist at commercial scale three years ago. The figures are extraordinary by any measure, and they describe a transformation that is still accelerating. The question is no longer whether artificial intelligence will become one of the defining industries of the twenty-first century — that question was answered when the revenue lines turned vertical. The question is whether the governance structures that humanity develops to manage this industry will prove adequate to the power of the technology it produces. The IPO, whenever it arrives, will not answer that question. But it will place the question before the largest and most consequential audience it has yet faced: the public markets, where the judgment of millions of investors will establish the price of artificial intelligence — and, in doing so, will signal what the world values more: capability or caution.