For the first time in recorded business history, the majority of companies are paying for artificial intelligence.
That's not a projection. That's what actually happened in March 2026, according to the Ramp AI Index - a monthly report built on real payment data from more than 50,000 businesses that use Ramp's corporate card and invoicing platform. The April update, published this week, put the number at 50.4%.
A year ago it was 35%.
That's a 15-percentage-point jump in 12 months. In business adoption cycles, that's not a trend - that's a shift.
What 50% Actually Means
There's a tendency to read a number like 50% and think "okay, half, that's about equal." But that framing misses what's happening underneath.
The businesses that adopted AI first are not at 50% AI usage. They're at 80%, 90%, fully integrated. The 50% figure is the midpoint of the overall population - which means the early adopters are far ahead of where the average suggests, and the laggards are further behind than 50% sounds.
Here's the more useful breakdown from the Ramp data:
By funding type:
- Venture capital-backed businesses: 80% AI adoption
- Private equity-backed businesses: 64% AI adoption
- Everyone else (bootstrapped, family-owned, small business): 45% AI adoption
That 35-point gap between VC-backed companies and everyone else is not a coincidence. And it's not going to close on its own.
Why VC-Backed Companies Are at 80%
Ramp's lead economist Ara Kharazian, who built this dataset, offers a clear explanation: venture capital firms are now functioning as an AI distribution channel.
When a VC firm invests in a portfolio company, they're increasingly bringing AI tooling with them - portfolio-wide deals, top-down directives to adopt the latest tools, and a cultural expectation that every team is experimenting with AI. The VC selection process now filters for AI-forward founders. The network effect does the rest.
Private equity shows a similar effect, just softer. PE firms tend to be less tech-forward culturally, so the push is real but weaker.
For the 45% of non-institutionally-backed businesses - which describes the overwhelming majority of small businesses in America - the adoption decision falls entirely to whoever is running the company. There's no VC mandate. No portfolio-wide tool rollout. No "AI champion" from the board. It's just you, figuring it out.
That's not an excuse to stay at 45%. It's context for why you might still be there.
The Anthropic Story Inside the Data
A secondary finding worth paying attention to: Anthropic is closing the gap with OpenAI at a rate that surprised even the Ramp economists.
In February, OpenAI held an 11-point lead in business adoption. By March, that gap had shrunk to 4.6 points - OpenAI at 35.2%, Anthropic at 30.6%. Anthropic added 6.3 percentage points in a single month.
Ramp's analysis notes that Anthropic already leads in three key sectors - information (tech/software companies), finance, and professional services. Those happen to be the three sectors with the highest overall AI adoption. The pattern in Ramp's historical data is clear: what the early adopters do today, the broader market does a few months later.
For a small business owner, this means the tool landscape is shifting. The AI you picked 18 months ago may not be the best tool today - and it may not be the dominant tool in your industry 18 months from now.
So What
Let's translate this into practical terms.
If your AI adoption is at zero: You're now in the 49.6% that hasn't paid for a single AI tool yet. That's fine, but your competitors - especially the ones with institutional backing or investor pressure - are not in that same position. The window to start without urgency is closing.
If you're using one or two AI tools casually: You're probably in the 50% club, but usage isn't the same as integration. Paying for ChatGPT is not the same as having AI embedded in how you actually run your business. The companies getting real gains have moved past casual use.
If you've been building AI into actual workflows: You're ahead. The 80% adoption rate among VC-backed companies doesn't just mean they have accounts - it means they have systems. That's the goal.
The Ramp data does not tell you which tools to use or exactly what to do. But it does confirm something important: the question of whether to use AI has been answered by the market. The question now is how much ground you've already lost - and whether you're willing to close it.
Where the Ramp Data Comes From
The Ramp AI Index is built from actual corporate card and invoice payment data across Ramp's platform. It tracks what businesses are paying for, not what they say they plan to use. That makes it one of the more reliable adoption datasets available - there's no self-reporting bias, no survey confusion. If a business paid for an AI tool, it shows up. If they didn't, it doesn't.
The full April 2026 update is available at ramp.com/leading-indicators/april-2026-ai-index.
Priya Kapoor covers the money side of running a small business: what the numbers say, what they don't, and what you should actually do about it.