Anthropic Overtakes OpenAI in US Business Adoption for the First Time

[Oscar_38]

The May 2026 Ramp AI Index dropped this week with a result that most people in AI were not expecting to see quite yet. Anthropic's Claude has overtaken ChatGPT in US business adoption, measured by the corporate spend data that Ramp collects across more than 50,000 American businesses. Claude's share rose 3.8 percentage points in April to reach 34.4 percent. OpenAI's share fell 2.9 points to 32.3 percent. That is the first time Anthropic has led since the Ramp index began tracking this metric. A separate IDC survey paints a more cautious picture of Claude's enterprise reach and the two datasets do not fully reconcile, but the direction of travel in the Ramp data is clear. Businesses that were defaulting to ChatGPT are routing meaningful spend to Claude instead.

The context matters here. Anthropic has been aggressively targeting regulated industries including healthcare, finance and law with a safety-first positioning that resonates in sectors where hallucinations carry real legal and operational risk. Their enterprise revenue mix is around 80 percent of total revenue versus OpenAI's roughly 40 percent. They are also projecting break-even two years ahead of OpenAI. The timing is interesting given the Fable and Mythos shutdown the same week, which could have the paradoxical effect of reinforcing Anthropic's safety-focused brand with enterprise buyers even as it disrupts the product offering. A company that complies with government directives immediately and maintains data retention policies specifically to catch jailbreaks looks like a reliable partner to a regulated industry procurement team.

Is the Anthropic vs OpenAI race more interesting to you as a technology competition or a business model competition? And which do you think is better positioned heading into the IPO window?

[Always_David72]

Spend data from a corporate card platform is probably the most honest measure of enterprise AI adoption that exists. It cannot be gamed by prompt count or API call volume. Real money moving is real adoption

[PlanetOftheApes]

The IDC survey disagreeing with Ramp is interesting. Those two sources measure different things. Ramp sees where finance approved spend goes. IDC surveys what IT managers say they are using. The gap between those is often significant

[Darren51]

Anthropic's safety positioning looking better after the Fable shutdown is genuinely counterintuitive but I think it is correct. Enterprises do not want a partner that fights regulators, they want one that complies and moves on

[Stephen24]

OpenAI having 900 million monthly active users versus Anthropic's 134 million is an enormous consumer gap. But if the IPO story is about enterprise revenue the consumer number almost does not matter

[Maxximus]

The 80 percent enterprise revenue number for Anthropic is the most important figure in all of this. It implies much more predictable revenue and much better margins than a consumer-heavy model

[Gareth5]

I want to know what the churn rates look like. Switching AI providers in enterprise is annoying but not impossible. If Claude's advantage is real it needs to be sticky over multiple contract cycles

[Leo]

OpenAI has Microsoft in their corner with Azure integration that is baked into enterprise software stacks. That distribution advantage does not show up in the Ramp data cleanly

[Lynx]

The healthcare and legal sector focus from Anthropic is smart because those buyers have compliance requirements that naturally push them toward the more cautious AI provider. It is a defensible niche

[TheRock25]

Two years ahead on break-even is going to attract a completely different type of public market investor than OpenAI will attract. Growth investors versus value-adjacent institutional money

[Seb93]

If I had to bet on one company still being dominant in enterprise AI in five years I would want to see the actual S-1 financials before answering. The narrative is compelling on both sides but the numbers will tell a different story