OpenAI Tightens Its Grip on Enterprise AI with 36.8%  Of Subscriptions as Corporate Spending Accelerates

After months of debate over whether OpenAI was losing ground to fast-rising rivals, new corporate spending data has suggested the opposite, with the ChatGPT maker noted tightening its grip on enterprise AI just as businesses move from experimentation to routine, budgeted use.

Fresh figures from Ramp, a U.S. startup that tracks corporate card and bill payments for more than 50,000 businesses, show OpenAI posting its strongest business growth in months in December 2025. The data captures billions of dollars in AI-related spending and offers a window into which companies are actually getting paid as AI becomes embedded in everyday corporate work.

The share of U.S. businesses paying for AI products and services rose to 46.6% in December, up 1.6 percentage points from November. That was the biggest month-over-month increase since mid-2025, signaling a renewed acceleration in enterprise adoption after a softer stretch earlier in the year.

OpenAI drove much of that growth. Ramp’s data shows that 36.8% of U.S. businesses on its platform were paying for OpenAI products in December, up two percentage points in a single month and a new record high. The rebound followed a brief slowdown in the fall, when concerns grew that Google’s Gemini and Anthropic’s Claude were starting to close the gap.

What stands out is not just the number of companies paying OpenAI, but how they are using it. Ramp’s line-item data shows gains across both enterprise chat subscriptions and API spending. That suggests OpenAI is being adopted across multiple layers of organizations, from non-technical staff using chat tools for writing, research, and analysis, to engineering and data teams integrating models directly into products, internal tools, and customer-facing systems.

This breadth matters as API usage, in particular, tends to signal deeper integration and longer-term commitment, because it ties AI models into workflows that are costly and disruptive to replace. It also points to OpenAI’s success in positioning itself as infrastructure rather than just a productivity add-on.

OpenAI does not disclose detailed revenue or user breakdowns, but the company said in November 2025 that it had reached 1 million business customers. The December Ramp data suggests those customers are not only sticking around but increasing spend, reinforcing the idea that OpenAI is benefiting from companies standardizing on a single AI platform as they scale usage across teams.

The broader context helps explain why. Many companies appear to be moving past trial phases and into recurring, operational use of AI for software development, customer support, finance, sales, and internal research. December’s jump, Ramp notes, reflects recurring spend rather than one-off experiments, a sign that AI is becoming part of routine operating budgets.

Competitors are still growing, though more unevenly. Anthropic’s adoption rose to 16.7%, with Ramp highlighting that growth was concentrated among technology companies making heavy use of APIs. That points to strong traction with developers and AI-native firms, but a narrower footprint across non-technical business functions.

Google’s AI adoption rose to 4.3%. Ramp cautioned that this likely understates Gemini’s reach because many companies access it through existing Google Workspace subscriptions at no incremental cost, meaning no separate payment shows up in transaction data. Even so, the numbers underscore a challenge for Google: widespread access does not automatically translate into visible, discretionary AI spending.

Ramp’s data also comes with clear limitations. It excludes free AI tools, bundled offerings, and cases where employees use personal AI accounts for work tasks. That means overall AI usage across U.S. businesses is almost certainly higher than the figures suggest. What the data captures, however, is paid, trackable commitment — the point at which AI moves from curiosity to line item.

OpenAI’s renewed momentum comes amid heightened competition, heavier scrutiny of AI costs, and growing debate over whether the rapid expansion of enterprise AI is sustainable. OpenAI CEO Sam Altman has acknowledged rising compute costs, while rivals have positioned themselves as more efficient or more specialized alternatives.

Yet the Ramp data suggests that, for now, OpenAI’s combination of brand recognition, model performance, developer ecosystem, and enterprise packaging is resonating with buyers. Once companies commit to AI at scale, switching costs rise, favoring incumbents that can serve multiple use cases under one contract.

In that sense, December’s rebound may say less about short-term feature battles and more about a structural shift. As AI becomes infrastructure, enterprises appear to be consolidating around providers they view as reliable, extensible, and broadly applicable. The numbers hint that OpenAI’s early lead is translating into a durable advantage, even as competitors continue to innovate and chip away at specific niches.

Now, it is becoming clear that winning mindshare is no longer enough for the rest of the market. The real contest is over who captures sustained enterprise spend as AI becomes a permanent fixture of corporate life.