A new analysis of corporate spending data indicates that Anthropic now holds a larger share of the business market for generative artificial intelligence services than OpenAI, marking a notable shift in the competitive landscape for enterprise AI tools.
The findings come from a survey compiled by the financial technology firm Ramp, which examined expense data from its client base. The data, released in early 2025, shows that 34.4 percent of participating businesses are paying for services from Anthropic, the developer of the Claude large language model. In comparison, 32.3 percent of those same businesses reported paying for services from OpenAI, the creator of ChatGPT.
The survey did not include every business in Ramp’s customer portfolio but was drawn from a sample of companies that use Ramp’s expense management platform. Ramp is a fintech company that provides corporate credit cards and spending management software to thousands of businesses, primarily in the United States.
Enterprise adoption trends
Anthropic was founded in 2021 by former OpenAI employees including Dario Amodei and Daniela Amodei. The company has focused heavily on safety research and enterprise deployment, releasing Claude 3 models in March 2024 followed by Claude 3.5 in June 2024. OpenAI, founded in 2015, has long been considered the market leader in generative AI, particularly following the viral adoption of ChatGPT in late 2022.
While OpenAI still maintains a larger overall user base due to widespread consumer adoption of ChatGPT, the Ramp data suggests Anthropic has gained meaningful ground in the business segment. The findings point to a narrowing gap in corporate adoption, driven by factors including pricing, model performance, and data handling policies.
Industry analysts have noted that enterprises often prioritize factors such as security, compliance, and predictable pricing when selecting AI providers. Anthropic’s contract structure and emphasis on “responsible AI” have resonated with certain corporate clients, according to multiple reports.
Methodology and context
Ramp’s survey data is not a comprehensive census of all enterprise AI spending but rather a snapshot of spending patterns observed among its own customers. The sample includes thousands of businesses, predominantly startups and midsize companies, which may skew toward tech and financial services sectors. The company stated that it aggregated anonymized expense data to produce the figures.
Neither Anthropic nor OpenAI has publicly confirmed the specific proportions of their commercial customer bases, nor have they commented directly on the Ramp findings. Both companies continue to release new models and compete for enterprise contracts, with major clients including financial institutions, technology companies, and healthcare organizations.
OpenAI reported in November 2024 that it had passed 200 million weekly active users for ChatGPT, but the company has not disclosed a breakdown between consumer and business users. Anthropic has similarly reported strong revenue growth but has not provided exact customer counts.
Broader implications for the AI market
The Ramp data arrives amid a broader surge in enterprise AI adoption, as companies across industries experiment with generative AI for tasks such as customer support, code generation, and content creation. The market for these services is projected to grow substantially over the next five years, attracting competition from additional players including Google’s Gemini models and Meta’s open source Llama series.
Spending on AI platforms remains a significant and increasing line item for many businesses, with some companies dedicating entire departments to integrating large language models into internal workflows. The Ramp data, while limited in scope, offers one of the few third party indicators of actual payment behavior among business users.
It remains to be seen whether Anthropic can sustain its lead in this segment as OpenAI continues to roll out enterprise features, including data privacy options and dedicated private deployments. Both companies are expected to release new model versions in the coming months, potentially influencing future purchasing decisions.
Source: GeekWire
