Most EU Business AI Is Agentic, OpenAI Says. How Much Does That Tell Us?

🔍 Read the full analysis: Most EU Business AI Is Agentic, OpenAI Says. How Much Does That Tell Us? on ThorstenMeyerAI.com

Buying for a business?Offer from Amazon

Get business pricing on tech for your team

• Business-only prices and quantity discounts
• Tax-exempt purchasing
• Multiple users, one account, clear invoices

Create a free Business account →Free to register

As an affiliate, we earn on qualifying purchases.Hide

OpenAI says its agent products generated 55% of output tokens in ChatGPT Enterprise and Business accounts across the 27 EU member states by September 30, 2026. The figure, published in a new company report, describes activity among OpenAI’s customers—not the share of European businesses’ AI tasks or economic value handled by agents.

OpenAI’s October 14 report, Europe’s Agentic Opportunity: Moving from breadth to depth with AI agents, counts output from Codex, the company’s coding agent, and ChatGPT Work, an agent mode launched in July that can work with files, browsers and company tools. Codex contributed 34% of the measured output tokens and ChatGPT Work 21%; ordinary ChatGPT accounted for the remaining 45%.

The report also says that among businesses with 200 or fewer users, weekly active users of Codex or ChatGPT Work grew 8.1 times between the first week of May and the week ending October 5. OpenAI says every EU member state recorded at least fivefold growth. Belgium, Italy and Luxembourg had the highest reported growth, at 12.1, 10.8 and 10.0 times, respectively. The company does not provide absolute user counts in the material cited.

OpenAI’s measure of “frontier firms” means the top 10% of companies by output tokens per active user. In September, EU frontier firms generated about four times as many tokens per user as typical EU firms, while frontier firms outside the EU generated more than twice as many as the EU frontier. OpenAI estimates that EU companies make up about 10% of frontier firms worldwide, compared with about 65% in North America.

What the 55% Figure Measures

The report adds detail about how OpenAI customers use its products and points to fast growth in agent activity on its platform. But tokens measure generated text, not completed tasks, the number of people using agents, or the value businesses get from them. OpenAI’s own Enterprise Signals page describes tokens as an imperfect measure of business value and says agentic tasks can require substantially more computation.

That distinction matters because an agent may read files, plan, write, check its work and try again, generating many more tokens than a short chat exchange. A larger token share can therefore reflect the demands of agent workflows as well as the prevalence of those workflows. The figure supports a narrower conclusion: in these accounts, agents produced more of OpenAI’s measured output than standard chat did.

The findings are also relevant to claims about Europe’s position in AI adoption. OpenAI argues that Europe needs to move from broad experimentation to deeper use, with more worker skills and confidence and greater access to advanced capabilities. Its customer data can inform that discussion, but it cannot establish how representative its customers are of European companies overall or whether greater token use translates into better business outcomes.

A Product Launch Shapes the Trend

The reported growth spans a period when OpenAI added a new product. According to the report’s chart, Codex accounted for roughly a fifth of EU output tokens in May and 34% by September 30. ChatGPT Work launched in July and accounted for 21% by the September endpoint. OpenAI says a major part of the small-business growth followed that launch.

This makes the timing important when interpreting the 8.1-fold figure. Its baseline is the first week of May, before ChatGPT Work existed. The increase shows that weekly active use of the included tools rose from that baseline in OpenAI’s smaller-business accounts. It does not, on its own, show that agent use grew at the same rate across Europe’s small-business sector, or that the change will persist.

The report’s country figures also vary widely: agentic output shares range from 85% in Lithuania and 81% in Cyprus to 22% in Slovenia and 26% in Austria. Germany is reported at 51%. OpenAI notes that smaller countries can have relatively high agentic use within larger firms. Without country-level account or user counts, those percentages should be read as a snapshot of OpenAI’s customer mix rather than a ranking of national AI readiness.

Limits of the Customer Data

The report covers OpenAI’s EU Enterprise and Business accounts; it does not provide a census of AI use among European businesses. The published figures also do not give absolute account or user counts for each country, making it difficult to judge how much particular customers influence national percentages. The source does not establish that token growth corresponds to more tasks completed, higher productivity or improved business performance.

It is also unclear how much of the reported adoption will continue after the initial period for ChatGPT Work, or how use compares with competing AI tools. The report’s country rankings and small-business growth figures describe different measures and do not by themselves show which countries are ahead in economy-wide adoption. OpenAI presents the findings as a view of its customer base, not a comprehensive assessment of the EU market.

Whether Deeper Use Delivers

OpenAI’s report calls for European businesses and policymakers to build worker skills and confidence and make advanced AI capabilities more accessible and affordable. The next useful evidence would show whether agent use persists beyond the launch period, how many organizations and workers are involved, and whether agents improve measurable outcomes such as time, cost or quality of work.

Until comparable data with clear baselines and business outcomes are available, the 55% figure is best treated as a measure of OpenAI platform output. It identifies a shift in the company’s own EU business accounts, while leaving the scale and economic effect of agent adoption across Europe unresolved.

Key Questions

What does OpenAI mean by “55% of output”?

It means Codex and ChatGPT Work produced 55% of the output tokens generated in the covered EU ChatGPT Enterprise and Business accounts by September 30, 2026. It is not a measure of 55% of users, tasks or business value.

Which tools does OpenAI classify as agentic?

The report includes Codex, OpenAI’s coding agent, and ChatGPT Work, its agent mode for working with files, browsers and company tools. Standard ChatGPT is counted separately.

Does the report show that most European companies use AI agents?

No. The figures cover OpenAI’s business accounts in the EU, not all European companies or AI products. The report does not establish what share of businesses or business tasks use agents.

Why might agent use produce more tokens than chat?

Agent workflows may involve reading material, planning, producing output and checking or revising it. Those steps can generate more tokens per workflow than a brief chat reply, so token share is not the same as the share of interactions.

What does OpenAI say Europe should do?

The report says Europe should move from broad adoption toward deeper use, including building workers’ skills and confidence and improving access to advanced AI capabilities.

Source: ThorstenMeyerAI.com