Pull to refresh
Logo
Daily Brief
Following Why Ranks Sign Up
Paid enterprise AI subscriptions cross 50 percent in the US

Paid enterprise AI subscriptions cross 50 percent in the US

New Capabilities

Ramp's March 2026 index records the first month more than half of US businesses paid for AI tools, driven by manufacturing and retail joining tech and finance

April 14th, 2026: Anthropic overtakes OpenAI in monthly Ramp data

Overview

ChatGPT launched in November 2022. In March 2026, 50.4 percent of US businesses paid for AI subscriptions, the first month adoption crossed half, according to Ramp's monthly index of corporate card spending. The reading was 46.8 percent in January and about 35 percent a year earlier.

The jump came from manufacturing and retail, sectors that had lagged tech and finance. The shift moves AI past the pilot phase into committed budgets. Hyperscalers now project $700 billion in compute spending this year, and Anthropic's enterprise share rose to 30.6 percent against OpenAI's 35.2 percent.

Why it matters

Half of US businesses now pay for AI tools, the threshold past which a technology starts to reshape labor demand, software vendor power, and capital spending.

Play on this story Voices Debate Predict

Key Indicators

50.4%
US businesses paying for AI (Mar 2026)
First month above half in Ramp's monthly index of corporate card spending.
+15 pts
Adoption growth in 12 months
Up from about 35 percent in March 2025; 46.8 percent in January 2026.
$2.5B
Claude Code annualized revenue
Anthropic's coding tool reached this run-rate within 10 months of launch.
$700B
Projected 2026 hyperscaler capex
Combined data-center and chip spending by AWS, Microsoft, Google, and Meta.
35.2% vs 30.6%
OpenAI vs Anthropic enterprise share
OpenAI plateaued in March; Anthropic gained six points in one month.

Interactive

Exploring all sides of a story is often best achieved with Play.

Ever wondered what historical figures would say about today's headlines?

Sign up to generate historical perspectives on this story.

Log in Sign Up

People Involved

Sam Altman
Sam Altman
Leading the AI vendor with the largest enterprise share, now at 35.2 percent and flat for the first time since 2023
 Dario Amodei
Dario Amodei
Running the AI vendor whose enterprise share rose from 24.4 to 30.6 percent in a single month
 Eric Glyman
Eric Glyman
Publishes the monthly index that first recorded the 50 percent crossing

Organizations Involved

Ramp
Ramp
Corporate fintech
Source of the monthly index that recorded the 50 percent crossing

A corporate card and spend-management platform whose anonymized customer data feeds the Ramp AI Index.
OpenAI
OpenAI
AI model developer
Largest US enterprise AI vendor by share; first month-over-month flat reading in March 2026

The AI lab behind ChatGPT and the GPT model family; the first vendor to reach mass enterprise distribution.
Anthropic
Anthropic
AI model developer
Second-largest US enterprise AI vendor; closed gap with OpenAI to under five points in March 2026

The AI lab behind Claude, currently the fastest-growing enterprise AI vendor on Ramp's data.
U.S. Census Bureau
U.S. Census Bureau
Federal statistical agency
Runs the Business Trends and Outlook Survey, which uses a narrower definition and shows about 18 percent adoption

The federal agency that runs the Business Trends and Outlook Survey (BTOS), a biweekly survey of US firms.

Timeline

  1. Anthropic overtakes OpenAI in monthly Ramp data

    Market Shift

    Anthropic's share of paying US businesses rises to 34.4 percent against OpenAI's 32.9 percent.

  2. Federal Reserve publishes AI adoption monitoring note

    Policy

    Fed economists publish a FEDS Note tracking AI adoption across surveys, citing wide variance in measured rates.

  3. Business AI adoption crosses 50 percent

    Milestone

    Ramp's March 2026 reading shows 50.4 percent of US businesses pay for AI tools, the first month above half.

  4. Ramp AI Index: 46.8 percent paid adoption

    Data Release

    Ramp's January update shows US business AI adoption at 46.8 percent, the highest reading on record.

  5. Menlo Ventures: Anthropic at 40 percent of enterprise LLM spend

    Market Report

    Menlo's annual report shows Anthropic at 40 percent of enterprise model spend, up from 12 percent in 2023.

  6. Claude Code hits $1B annualized run-rate

    Milestone

    Anthropic reports Claude Code reached $1 billion in annualized revenue within six months of launch.

  7. Claude Code launches

    Product Launch

    Anthropic releases Claude Code, a terminal-based developer agent. It becomes the company's fastest-selling product.

  8. ChatGPT Enterprise released

    Product Launch

    OpenAI launches a business tier with data controls. The product anchors AI in enterprise budgets.

  9. GPT-4 launches; enterprise pilots begin

    Product Launch

    OpenAI ships GPT-4 with API access. Fortune 500 firms begin paid pilots through Microsoft Azure.

  10. ChatGPT launches

    Product Launch

    OpenAI releases ChatGPT as a free web preview. The product reaches 100 million users in two months.

Scenarios

Predict which scenario wins. Contrarian picks score more — points lock in when the scenario resolves.

Log in to predict. Track your picks, climb the leaderboard. Log in Sign Up
1

Paid AI adoption climbs to 70 percent of US businesses by year-end 2026

Manufacturing and retail led the March jump. If small business follows the same pattern with a six- to nine-month lag, the index keeps climbing through 2026. Ramp's own commentary points to compounding agent deployments as a driver: once a firm pays for one AI agent, it tends to pay for more.

Resolves by: 2026-12-31
Source: Ramp AI Index December 2026 monthly update
Discussed by: Ramp researchers in the March 2026 update; VentureBeat coverage of the index
Consensus —
2

Anthropic holds the top US enterprise share for three straight months

Anthropic gained six points in March and overtook OpenAI in April on the same index. The question is whether Claude Code's developer share carries over to broader enterprise spend. Three consecutive months on top would mark a durable lead rather than a single-month swap.

Resolves by: 2026-09-30
Source: Ramp AI Index monthly vendor share readings (July, August, September 2026)
Discussed by: Axios, The Decoder, and VentureBeat, citing the Ramp series and Menlo Ventures data
Consensus —
3

AI productivity gains show up in BLS data by mid-2027

Historical general-purpose technologies took years to register in productivity statistics. The Fed note warns measured AI adoption varies widely. If the Ramp 50 percent reading reflects real workflow change, US nonfarm business productivity should run above its 2010s trend for at least two consecutive quarters in 2027.

Resolves by: 2027-09-01
Source: US Bureau of Labor Statistics quarterly nonfarm business sector labor productivity release
Discussed by: Federal Reserve FEDS Notes (April 2026); Brookings AI productivity tracking
Consensus —
4

Adoption stalls below 60 percent through year-end 2026

Census data shows broad adoption is much lower than Ramp's paying-customer figure. Many firms outside Ramp's customer base, including the smallest businesses and regulated industries, may take longer. If the next wave of buyers is slower to commit, the curve flattens in the high 50s.

Resolves by: 2026-12-31
Source: Ramp AI Index December 2026 monthly update
Discussed by: Census Bureau BTOS researchers; small-business analysts citing the SBA Advocacy report
Consensus —
5

A US federal AI procurement or disclosure rule lands in 2026

A majority of businesses paying for AI changes the political math. Procurement standards, model disclosure rules, or audit requirements become easier to enforce when adoption is widespread. The Office of Management and Budget or Federal Trade Commission could issue binding rules before year-end.

Resolves by: 2026-12-31
Source: Federal Register
Discussed by: Reuters, Politico coverage of the FTC and OMB AI guidance
Consensus —

Historical Context

Factory electrification in the US (1900s-1920s)

1900-1929

What Happened

After Edison opened the first commercial power station in 1882, US factories took two decades to shift from steam to electric drive. By 1919 about half of US manufacturing capacity ran on electricity. The transition required rewiring buildings, retraining workers, and rethinking factory layouts.

Outcome

Short Term

Manufacturing output per worker began rising in the 1920s. Capital spending on electric motors and generators surged.

Long Term

Productivity gains from electrification did not show in aggregate statistics until the late 1920s, roughly a decade after the 50 percent threshold. Economist Paul David documented this lag as a recurring pattern for general-purpose technologies.

Why It's Relevant Today

Electrification's adoption curve is the closest historical match for AI: a general-purpose technology with a long gap between deployment and measured productivity. The Federal Reserve cited this analogy in its April 2026 note on AI.

World Wide Web crosses 50 percent of US households

2000-2001

What Happened

Tim Berners-Lee released the Web in 1989. Netscape's browser shipped in 1994. By 2000 about 50 percent of US households had internet access, mostly dial-up. Business adoption ran ahead of households; most large firms had public websites by 1998.

Outcome

Short Term

The dot-com crash wiped out roughly $5 trillion in equity value between March 2000 and October 2002. Surviving firms consolidated.

Long Term

E-commerce, cloud software, and digital advertising grew into the dominant business categories of the 2010s. Productivity gains showed clearly in US data from 1996 to 2004, then slowed.

Why It's Relevant Today

The Web reached 50 percent household penetration roughly seven years after Mosaic shipped. AI reached 50 percent business adoption in about 40 months from ChatGPT, the fastest enterprise technology diffusion on record.

Cloud computing crosses 50 percent of enterprise IT spend

2018-2020

What Happened

Amazon Web Services launched in 2006. By 2018, Gartner reported public cloud spending had passed half of new enterprise IT infrastructure budgets. The shift was driven first by startups, then by Fortune 500 migrations led by AWS, Microsoft Azure, and Google Cloud.

Outcome

Short Term

On-premise hardware revenue fell at HP, Dell, and IBM. Cloud vendors became the largest software businesses by revenue.

Long Term

Public cloud reshaped the enterprise software vendor stack and concentrated infrastructure spend on three providers. The same providers are now the largest buyers of AI compute.

Why It's Relevant Today

Cloud adoption shows how a 50 percent enterprise threshold reshapes vendor economics within five years. The same hyperscalers that won cloud now sit at the center of AI capex, which is projected to hit $700 billion in 2026.

Sources

(10)
+4
Human Progress Ramp FinancialContent Axios The Decoder VentureBeat Federal Reserve Board US Census Bureau Menlo Ventures Ramp