OpenAI tilts to enterprise

OpenAI built its fame on one person typing into ChatGPT. By April 2026, its chief revenue officer said companies now generate more than 40% of OpenAI’s revenue, and OpenAI said that share could reach parity with consumer revenue by the end of 2026. (openai.com) (cnbc.com) That changes what buyers are paying for. A consumer will tolerate a clever answer in a chat box, but a bank, insurer, or drug company wants software that can plug into approvals, logs, identity systems, and contracts. (openai.com) (finance.yahoo.com) OpenAI is now selling that shift as “agentic workflows,” which means the model does not just answer a question once. It takes a task like writing code, searching documents, or updating a system and handles several steps in sequence for a team. (openai.com) (decrypt.co) The company’s own numbers show why that pitch is landing. OpenAI said its application programming interfaces now process more than 15 billion tokens per minute, and its Codex coding product has reached 3 million weekly active users. (openai.com) Those are not hobbyist names on the customer list. OpenAI said new enterprise customers include Goldman Sachs, Philips, and State Farm, while existing business users include DoorDash, Thermo Fisher, Cursor, and LY Corporation. (openai.com) Once artificial intelligence moves from a single employee’s browser tab into a company’s internal systems, the buying checklist gets longer. OpenAI’s enterprise policy says business customers get data ownership, no training on their data by default, retention controls, single sign-on through Security Assertion Markup Language, and compliance documentation such as System and Organization Controls 2. (openai.com) That is also where the competition has moved. SiliconANGLE reported on April 9 that Anthropic rolled out organization-wide controls for its Claude Cowork service while OpenAI cut pricing on its Pro plan for Codex access, which shows both companies are fighting on procurement terms as much as on model quality. (siliconangle.com) OpenAI’s leadership is talking about this like a company preparing for public-market scrutiny, not like a lab chasing viral demos. CNBC reported on April 8 that chief financial officer Sarah Friar said a company at OpenAI’s scale should “look and feel and act” like a public company, while Denise Dresser tied the compute buildout directly to enterprise demand. (cnbc.com) The practical result is that artificial intelligence vendors are starting to look more like old enterprise software companies. The new sales pitch is less “try this chatbot” and more “we can give your legal team logs, your information technology team access controls, and your procurement team a contract they can sign.” (openai.com) (siliconangle.com) If that mix keeps moving toward business buyers, OpenAI’s center of gravity shifts with it. The product decisions that matter most will be the ones that survive a security review, fit into a budget line, and save a large company enough labor hours to renew next year. (openai.com) (finance.yahoo.com)