• OpenAI's business-to-business revenue more than doubled from the start of Q3 through late September, according to people familiar with the matter.
  • The company's annualized revenue run rate rose over 70% to nearly $70 billion, though the figure remains unaudited and private.
  • The enterprise push, including ChatGPT Business and coding tools, signals a strategic shift toward stickier corporate contracts.

OpenAI's efforts to build a large enterprise software business are gaining traction. Business-to-business revenue more than doubled from the beginning of the third quarter through late September, according to a person familiar with the company's finances. The surge helped lift OpenAI's annualized revenue run rate—a forward-looking measure of recurring sales—by more than 70% to nearly $70 billion. The figures are reported, not drawn from a public, audited quarterly filing; OpenAI remains privately held.

The acceleration points to a broadening of OpenAI's commercial base beyond consumer subscriptions. The company now markets ChatGPT Business and Enterprise, developer APIs, coding products such as Codex, and enterprise agents that can operate within approved company applications, files, and processes. Recent updates to its Business plan include templates that connect to GitHub Enterprise, Snowflake (SNOW), and Databricks—integrations designed to embed AI more deeply into corporate workflows and make subscriptions harder to cancel. According to Axios, which first reported the figures, consumer revenue added in Q3 alone exceeded all consumer revenue generated in 2025.

A Push Beyond the Chatbot

While ChatGPT remains the public face of OpenAI, the company's commercial story is increasingly enterprise- and agent-focused. That shift matters because enterprise contracts tend to be recurring, can justify investment in reliability and governance, and offer a path toward more predictable revenue than one-off consumer purchases. The strategy also aligns with the demands of large, regulated customers: OpenAI's Enterprise offering emphasizes administrative controls, auditability, and the ability to use approved internal tools and data.

The growth comes amid a broader move from generative-AI experimentation to deployment in core business processes. Companies are buying AI for programming, internal knowledge retrieval, document work, data analysis, and customer operations, as well as for semi-autonomous agent workflows. The gains create demand well beyond software—for data-center construction, electricity generation, networking, advanced chips, and cloud services. OpenAI says its Stargate initiative, a joint project with SoftBank (9984.T) and Oracle (ORCL) announced as a potential investment of up to $500 billion over four years, has surpassed its initial 10-gigawatt U.S. infrastructure goal, with more than 3 GW of added capacity in the preceding 90 days.

Profitability Still an Open Question

Rapid sales growth does not settle the profitability question. OpenAI faces exceptionally large costs for model training, inference, chips, cloud services, and data centers. Reuters has reported internal projections for spending through 2030 reaching $750 billion. The company closed a March funding round with $122 billion in committed capital at an $852 billion valuation, and later held discussions on financing that could imply roughly $1.2 trillion before an IPO, according to Reuters. Those are reported private-market figures, not public-market prices.

In October 2025, OpenAI recast its for-profit subsidiary as OpenAI Group PBC, a public-benefit corporation controlled by the OpenAI Foundation. Microsoft (MSFT) retained an approximately 27% as-converted stake, valued by Microsoft at about $135 billion. The restructuring was designed to make fundraising and operations more conventional while preserving nonprofit control and the strategic relationship with Microsoft, which retains key rights and Azure-related exclusivity through the partnership framework. OpenAI has gained latitude to develop certain products with third parties.

Governance and Policy Risks

OpenAI's governance evolution has been unusually consequential. The November 2023 removal and rapid reinstatement of CEO Sam Altman exposed tensions between the organization's original nonprofit mission, safety governance, and the commercial demands of leading the generative-AI market. Subsequent departures included co-founder and chief scientist Ilya Sutskever in 2024 and several other senior technical leaders.

Policy is now central to the enterprise-AI market because major customers need confidence that deployments are lawful, auditable, and safe. In the European Union, the AI Act is the first comprehensive AI legal framework. Its transparency rules took effect in August 2026, while high-risk-system obligations phase in later. OpenAI has published EU-focused customer guidance and states that customers must avoid prohibited uses, including certain manipulative practices, social scoring, some biometric uses, and emotion inference in workplaces or schools outside narrow exceptions.

In the United States, copyright and competition disputes remain material risks. The New York Times (NYT) and other publishers allege that OpenAI and Microsoft used copyrighted journalism without permission for training; the Trump administration filed a September brief supporting the companies' fair-use position. OpenAI is also contesting claims connected to the Apple (AAPL)–OpenAI integration. xAI and Elon Musk's companies alleged that the partnership unlawfully excluded rivals; OpenAI has asked a federal court to dismiss the suit.

The reported revenue surge provides evidence that OpenAI may be building a large recurring software business capable of helping finance its infrastructure ambitions. But the key economic question remains whether gross margins and cash flow can eventually support the scale of compute spending required to compete at the frontier. For now, the enterprise momentum is real—and so are the costs and controversies that come with it.