Industry

AI-generated text

OpenAI Narrows Gap on Anthropic Among U.S. Corporate Users, Ramp Data Shows

New data from corporate card and expense manager Ramp indicates OpenAI is closing the gap with Anthropic among U.S.

OpenAI Narrows Gap on Anthropic Among U.S. Corporate Users, Ramp Data Shows

With OpenAI and Anthropic not yet near the point of releasing IPO financials, observers must rely on alternative datasets to assess how their businesses are performing. One such dataset comes from Ramp, a corporate credit card and expense management company, which has published fresh figures showing OpenAI gaining ground on Anthropic among U.S. business customers.

According to Ramp, OpenAI — once the clear leader across both consumer and business segments — lost its lead among Ramp’s paying business users in May, when Anthropic reached a 41% share versus OpenAI’s 39%. OpenAI has not reclaimed that top spot. By July, Anthropic’s share was nearly 44% while OpenAI’s was nearly 40%.

Ramp’s data set covers more than 70,000 U.S. companies that process billions of dollars through Ramp’s bill pay and corporate card products. Ramp’s customers span multiple industries, though as a popular Silicon Valley corporate card provider the sample skews toward the technology sector.

Ramp economist Ara Kharazian says a closer look at the most recent data shows OpenAI is growing faster than Anthropic among this segment so far in Q3. There is still a month left in the quarter, however, so the trend could change before the period closes. Ramp declined to provide absolute dollar figures, sharing only percentage market shares.

This dataset is not a complete measure of the total market: it excludes large enterprises that use spend-management tools from providers such as American Express rather than Ramp. Still, the sample gives useful market signals. Those signals suggest Anthropic’s lead is not permanent: business customers appear willing to switch between providers as each lab releases new models, a churn that should concern investors assessing how “sticky” enterprise AI spending will be.

On X, Kharazian attributed some of OpenAI’s momentum to the new GPT-5.6 Sol, writing: “GPT-5.6 Sol is really good, increasingly the choice for developers.” He contrasted that with Anthropic’s higher-tier model: “Fable 5, meanwhile, disappointed both in adoption and real-world application given price + data retention requirements imposed by regulators.”

That characterization may be simplified. Fable — Anthropic’s premium model tier — carries a higher price and is aimed at more targeted use cases than a general chatbot. Nevertheless, Anthropic drew criticism when it informed Fable users that it must retain their data for 30 days.

Ramp’s figures also suggest that both companies should be able to grow business revenue even as they fight for share, because the overall market is expanding. Among Ramp customers the proportion of companies paying for AI exceeded 50% in March and climbed to nearly 56% by July.

In short: within Ramp’s sample Anthropic still leads among paying U.S. business users, but OpenAI’s faster Q3 growth has narrowed the gap. A fuller picture of the enterprise AI market will depend on broader datasets and the financial disclosures that may come with IPOs or other official reports.