Bay Street Wire
Tech & Business

Enterprise AI Market Volatility Underscores Lack of Vendor Stickiness

Portrait of Dev Okonkwo
Dev OkonkwoAI & machine learningAug 21AI
Enterprise AI Market Volatility Underscores Lack of Vendor Stickiness

AI-generated image · Bay Street Wire

New data from corporate spend management firm Ramp shows businesses rapidly switching between OpenAI and Anthropic as model performance and pricing shift.

Data from corporate credit card and expense management company Ramp, as first reported by TechCrunch, reveals a volatile battle for enterprise market share between OpenAI and Anthropic, suggesting that business loyalty in the AI sector is highly fluid.

According to reporting from TechCrunch, Anthropic took the lead among Ramp's paying business users in May, capturing 41% of the market compared to OpenAI's 39%. By July, Anthropic's share grew to nearly 44%, while OpenAI sat at nearly 40%. However, Ramp economist Ara Kharazian noted on X that OpenAI is currently growing faster in the third quarter, driven by the adoption of the GPT-5.6 Sol model among developers.

Conversely, Kharazian indicated that Anthropic's high-end Fable 5 model disappointed in both real-world application and adoption, citing price points and regulatory data retention requirements. TechCrunch notes that Anthropic sparked backlash after informing users of the Fable model that it is required to keep their data for a 30-day period.

While the total market is expanding—with the percentage of Ramp customers paying for AI rising from over 50% in March to nearly 56% by July—the rapid shifts in preference indicate a lack of "stickiness" in enterprise spending. Ramp's data, which covers more than 70,000 American businesses, suggests companies are willing to switch providers as different labs release new models.

Note that Ramp's data excludes large enterprises using spend-management tools from providers such as American Express, and the company declined to provide specific dollar figures spent.

Sources

More from Dev Okonkwo