OpenAI says agentic tools exceed half of output tokens in its EU business sample
NO IMAGE
AI Agents & Automation

OpenAI says agentic tools exceed half of output tokens in its EU business sample

OpenAIOct 14, 2026

OpenAI’s October 14 analysis finds agentic tool use rising among EU business customers, with substantial differences between countries and firm sizes.

OpenAI published an analysis of agentic AI use among businesses in the European Union on October 14, 2026, drawing on activity in its own ChatGPT Enterprise and Business customer accounts. The company says Codex and ChatGPT Work together account for more than half of output tokens in its EU enterprise sample, describing a shift toward tools that perform tasks rather than only answer prompts.

The report also tracks small and medium businesses, which OpenAI defines as organizations with 200 or fewer associated users. Since May 2026, weekly active users of Codex and ChatGPT Work at these EU businesses grew more than eightfold, according to the analysis, with growth reported in every member state. These are OpenAI product-use figures, not a survey of all European companies or a measure of region-wide AI adoption.

Use varies considerably across the enterprise accounts included. The share of output tokens attributed to Codex and ChatGPT Work ranged from 22% in Slovenia to 85% in Lithuania. OpenAI also found differences in the types of work users bring to ChatGPT across countries, though the post does not provide a detailed country-by-country breakdown of those tasks.

OpenAI compared its EU “frontier firms”—the top 10% of firms by output tokens per active user—with other businesses. In September, EU frontier firms generated more than four times as many output tokens per active user as typical EU firms. Yet firms outside the EU at the global frontier produced more than twice the per-user output of the EU frontier group. OpenAI says EU-based companies made up 10% of the global frontier-firm sample.

The figures indicate differences in intensity within the company’s customer base, but token volume alone does not establish productivity gains or business outcomes. OpenAI CFO Sarah Friar said broader access, worker skills and confidence would be important to expanding use. The analysis is best read as a snapshot of how customers of OpenAI’s business products are using its agentic tools, not as a representative census of the EU economy.

Industry Impact

Agentic AI use rises in OpenAI’s EU customer sample

EU business operations and enterprise software

Overview

OpenAI reports that Codex and ChatGPT Work account for more than half of output tokens in its EU enterprise sample, while weekly active users at its small and medium business customers grew more than eightfold since May. The figures suggest rising use of task-performing tools among OpenAI’s business customers, but they do not represent the wider EU economy or demonstrate productivity gains. Business leaders can use the findings as a prompt to assess adoption and outcomes within their own organizations; product managers may find the country and firm-size variation relevant when planning deployment.

Key Impact Points

Agentic tools account for over half of output tokens in OpenAI’s EU enterprise sample, indicating substantial use within that customer base.

Weekly active users of Codex and ChatGPT Work at OpenAI’s EU small and medium business customers grew more than eightfold since May 2026.

Adoption intensity varies widely across sampled countries, and token volume alone does not establish productivity or business outcomes.

Opportunities

Evaluate Agentic AI Adoption

EU business operations and enterprise software

The article does not establish a distinct commercial opportunity or show that higher token use produces better business results. It does point to a practical opportunity for organizations already considering agentic tools: identify suitable workflows, test them with users, and measure task outcomes. OpenAI CFO Sarah Friar highlighted broader access, worker skills, and confidence as important to expanding use. Businesses can use those factors to shape enablement and pilot plans, while comparing results across teams rather than relying on activity volume. Any service or product opportunity would need to demonstrate value through measured operational results.

Reference link

https://openai.com/index/the-eus-next-chapter-in-ai-is-agentic

Image Source Link

https://media.wired.com/photos/6aaade9cf251bfe9554bf39a/master/pass/OpenAI-Releases-New-AI-Policy-Business-2294958021.jpg

Community Discussion (0)

Loading discussions...