How did Anthropic do it?

Ramp AI Index shows overall business AI adoption rose to 47.6% of businesses in February, a record high, with 24.4% of businesses now using Anthropic. OpenAI’s adoption rate fell by 1.5%.

Anthropic adoption grew 4.9% month over month, its largest monthly gains since we started tracking. Nearly one in four businesses on Ramp now pays for Anthropic (a year ago, it was one in 25). OpenAI’s 1.5% decline was the largest in any single month for any AI model company since we started tracking business AI adoption. Google adoption grew slightly to 4.7% while xAI held at less than 2% of businesses.

I’ve long thought each AI company has its own distributional advantage. Google’s is that it already sells to most businesses — Gemini comes in the Google Workspace bundle for free. OpenAI's was that it was the consumer default; ChatGPT was where most people first encountered AI, and that consumer momentum carried into strong business adoption. Today, OpenAI is still the AI model company used by the most businesses.

Anthropic's distributional advantage was different. They were popular early on with the earliest adopters — the evangelists, the engineers, the “AI guy” on your team. And now Anthropic is leveraging that early-adopter base to go mainstream, with our data suggesting it’s becoming the consumer default in a way we used to talk about OpenAI.

And we see it in Ramp data: Anthropic now wins about 70% of head-to-head matchups against OpenAI among businesses purchasing AI services for the first time. It’s a complete reversal of the trends we observed in 2025, when OpenAI adoption accelerated faster than any other model company.

So how did Anthropic pull this off?

Early in the AI race, the conventional wisdom was that AI model companies couldn't build moats. A year ago, the debate was whether these companies could even be profitable as the cost of inference fell toward zero and models became commodities. Performance gaps between frontier models would shrink, usage would become near-free, and the whole thing would look like a race to the bottom.

That didn’t happen. Businesses are showing real preferences, and I don't think it's purely about performance.

Consider: Anthropic's Claude Code and OpenAI's Codex are roughly comparable products. On certain benchmarks, Codex is arguably better. Codex is also cheaper. And yet Anthropic can't meet its own demand. Every plan (consumer, pro, enterprise, API) still has usage limits and rate caps. Anthropic is actively turning away revenue because it doesn't have the compute to serve it. Despite charging more for roughly equivalent performance, Anthropic’s demand is growing.

That's unusual. In most enterprise markets, when two products perform similarly, the cheaper one often wins. Enterprises are not sentimental. They don't typically develop brand loyalty to software companies (vendor lock-in can look similar but is different). And yet here we are.

So what is the moat?

Maybe it's culture. Maybe it's that Anthropic has become — for lack of a better word — cool.

Katy Perry recently announced she's switched to Claude:

Brian Schatz, a prominent tech policy voice in the Senate, said that he's started using Claude.

Both came in the wake of the backlash over OpenAI's agreement with the Department of Defense, a moment that sharpened the public contrast between the two companies. Anthropic positioned itself differently, and a certain class of user noticed.

Is this the moat in AI? Not benchmarks and pricing but whether your model is the one that culturally aware people want to use? There might be a world in which choosing between OpenAI and Anthropic becomes less like an enterprise procurement decision and more like the green bubble/blue bubble distinction in iMessage: a signal of identity as much as a choice of technology.

That sounds absurd for an enterprise software category. But Ramp data suggests something like it is already happening.