What happened
Ramp released its latest AI Index, a data set that tracks paid AI product and service usage across more than 85,000 U.S. businesses using corporate card and bill‑pay transaction data. The index shows that 46.6% of firms recorded a payment for an AI tool in the most recent month, a figure far higher than the 20% adoption rate reported by the U.S. Census. ‑based spend rose 12.0% week‑over‑week and token volume grew 27.2% versus the same week a month earlier. OpenAI’s adoption slipped slightly (+0.3 pp) while Anthropic saw a 2.1‑point gain, reaching 14.3% of firms. The report also notes that firms adopting AI tend to add about 10% more employees, including entry‑level hires, according to a joint analysis with Revelio Labs.
Ramp’s AI Index aggregates anonymized corporate card, invoice and ACH transactions from its Bill Pay platform, identifying AI‑related spend through merchant names and line‑item descriptions. Firms are counted as AI adopters when they record a positive payment for an AI product or service in a given month.
The latest weekly snapshot shows total spending up 12.0% and token volume up 27.2% compared with the same week a month earlier. These metrics are derived from businesses that have linked their AI providers to Ramp’s AI Spend Management product, which imports usage and billing data from providers such as OpenAI, Anthropic, and major cloud platforms.
Adoption rates differ markedly by provider: OpenAI’s share of adopting firms rose marginally by 0.3 percentage points, while Anthropic’s adoption grew by 2.1 points to 14.3% of firms, marking its second‑largest monthly increase.
A joint analysis with Revelio Labs, which matches Ramp spend data to workforce records, finds that firms adopting AI experience a 10% increase in employment, including entry‑level hires, suggesting AI adoption is linked to broader hiring trends rather than pure automation displacement.
Ramp’s lead economist Ara Kharazian notes that the index likely underestimates true AI usage because many businesses employ free tools or personal accounts that leave no corporate transaction trail.
Why it matters
The Ramp AI Index provides a more transaction‑based view of AI uptake than traditional surveys, suggesting that corporate AI spending is more widespread than official statistics indicate. By capturing actual payments for AI services, the index reduces under‑reporting bias inherent in self‑reported surveys, offering investors, policymakers, and business leaders a clearer signal of how deeply AI is embedded in day‑to‑day operations. The rapid rise in spend and volume signals growing reliance on API‑driven AI models, which could pressure pricing, influence cloud provider negotiations, and shape future workforce planning as firms expand hiring to support AI‑driven initiatives. Moreover, the divergence between the Census figure (20%) and Ramp’s estimate (46.6%) highlights a data gap that could affect economic forecasting and regulatory assessments of AI’s macroeconomic impact.
The index’s transaction‑based methodology offers a more immediate and arguably more accurate gauge of AI penetration than traditional surveys, which can suffer from ambiguous questions and delayed reporting.
Higher spend and volume indicate that businesses are moving beyond exploratory pilots to sustained, production‑level consumption of AI models, which may affect cloud infrastructure demand and pricing negotiations.
The 46.6% adoption figure suggests that AI is becoming a standard business utility, potentially reshaping competitive dynamics across sectors as firms that lag in AI integration may face productivity disadvantages.
The observed employment boost challenges narratives that AI primarily displaces workers, instead pointing to a net hiring effect, at least in the short term, as firms expand teams to develop, manage, and integrate AI solutions.
The divergence between official Census data and Ramp’s estimate underscores a measurement gap that could influence policy decisions, such as funding for AI research, workforce development programs, and regulatory oversight.
Interactive Mechanism: How It Actually Works
Explore the underlying technology behind this development interactively.
A route planner searches possible journeys using explicit rules. What does this illustrate about AI?
What to watch next
Future releases of the Ramp AI Index will reveal whether the current growth in spend sustains, how adoption rates evolve across industry sectors, and whether pricing pressures emerge as more firms consume AI APIs. Analysts should monitor whether the gap between survey‑based adoption metrics and transaction‑based data narrows, which could signal improved measurement standards. Additionally, the competitive dynamics between AI providers—evidenced by Anthropic’s recent gains—may shift market share and influence pricing strategies, especially as enterprises seek cost‑effective alternatives to dominant platforms.
Subsequent weekly releases of the Ramp AI Index to track whether spend growth continues or plateaus, which would signal market saturation or pricing pressures.
Sector‑specific adoption trends that could reveal which industries are leading AI integration and which remain laggards, informing targeted investment or policy interventions.
Potential shifts in provider market share, especially if Anthropic’s growth continues, which could drive competitive pricing or new rollouts.
Changes in the gap between survey‑based adoption metrics and transaction‑based data, indicating whether broader measurement standards are being adopted by government agencies.
Any emerging patterns in free‑tool usage versus paid adoption, which could affect future revenue models for AI vendors and the overall economics of AI deployment.