What drives AI adoption?
Ramp AI Index April 2026 Update
Dear Colleagues: Today’s letter includes my monthly update of Ramp AI Index, our flagship research using spend data from Ramp to track how American businesses are using AI. Read the FT’s write up here.
Extras from Ramp data:
Trust issues: I went live with NYSE to talk about the trust and reliability issues keeping some firms from using AI. (youtube)
Field trip: Should we send Econ Lab intern Rafael Hajjar to the Strait of Hormuz? (twitter)
Ramp AI Index shows business AI adoption crossed 50% for the first time in March, reaching 50.4% of businesses. A year ago, it was 35%. Half of businesses on Ramp now pay for AI.
Anthropic continued its surge, growing from 24.4% to 30.6% of businesses — a 6.3 percentage point gain, surpassing last month’s record monthly gain. The gap between OpenAI (35.2%) and Anthropic is now just 4.6 points, down from 11 points in February. At this pace, Anthropic will overtake OpenAI within the next two months.
A note on our methodology: Ramp AI Index is developed using business spend data from Ramp. We count corporate card and invoiced-based payments in our measurements. Read about our research approach here.
How businesses learned to stop worrying and love Anthropic: Last month, the Defense Department designated Anthropic a security risk (a judge has temporarily blocked the designation, and reports suggest the Pentagon will let some teams use it anyway). And yet Anthropic’s adoption accelerated. Setting aside the legal question, the credibility of the government’s threat is significantly weakened by the fact that businesses shrugged it off.
So, it’s been a great run for Anthropic. In fact, Anthropic is already in the lead in three key sectors: information (software), finance, and professional services — the sectors with the highest AI adoption already. There’s a more interesting question embedded in this dataset: if half of businesses now pay for AI, what separates the half that does from the half that doesn’t?
Follow the money. When you ask “what kind of company is most likely to use AI,” the intuitive answers win: tech companies, big companies, companies in San Francisco. But the single strongest predictor of AI adoption in our data is who funded you.
Venture-capital-backed companies have an 80% AI adoption rate. Private-equity-backed companies: 64%. Everyone else: 45%. It holds in every single industry we track.
Even within the lowest-adoption industries, the VC effect holds. VC-backed construction firms (77%) and food companies (67%) adopt AI at rates well above their sector averages. The funding source predicts AI adoption better than the industry the business operates in.
Why does funding background affect AI adoption? Part of it is selection. VCs have always selected founders they believe they can add value to, historically through advice, introductions, and business partnerships. Today, that value proposition increasingly includes setting portfolio companies up with the right tooling and technology. And the selection runs both ways: what VC in 2026 is going to fund a founder who isn’t using AI?
I also think VCs are functioning as a transmission mechanism for AI adoption across their portfolios through top-down directives, portfolio-wide deals, and the general cultural expectation that you should be using the latest tools. These factors push adoption rates above what you’d see from organic discovery alone. PE firms have a version of this too, but the effect is weaker, in part because PE firms tend to be less tech-forward than their VC counterparts, and therefore not as far along on the AI adoption curve. At companies without institutional investors, the adoption decision is left to senior leaders and teams figuring it out on their own.
A prediction
Look to the early adopters. Among VC-backed firms, Anthropic (66%) already leads OpenAI (59%). It also leads in the three highest-adoption sectors: information (63% vs 54%), finance (52% vs 46%), and professional services (47% vs 44%). OpenAI still leads in every other sector. I don’t think that lasts. The pattern in our data is consistent: what early adopters do today, the broader market does a few months later. The early adopters are picking Anthropic.
Something from my camera roll
Standard economic intuition teaches that, in a loneliness epidemic, heart shaped boxes of chocolate will be marked 30% off.



Hi, Ara! Love your work I discovered tonight. How long did it take you to pick up on this stuff and start learning faster? What basic foundational knowledge do you need to analyze like this? Any books? Courses?