OpenAI’s EU Agentic AI Statistic Needs More Than A Headline

🔍 Read the full analysis: OpenAI’s EU Agentic AI Statistic Needs More Than A Headline 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 reported on Oct. 14 that its two agent products generated 55% of output tokens in its ChatGPT Enterprise and Business accounts across the 27 EU member states by Sept. 30. The figure signals a growing role for agents in OpenAI’s business products, but it measures generated text rather than the share of work tasks or business value handled by agents.

OpenAI Economic Research’s report, Europe’s Agentic Opportunity: Moving from Breadth to Depth with AI Agents, defines agentic use as activity through Codex, the company’s coding agent, and ChatGPT Work, its agent mode for working with files, browsers and company tools. Codex accounted for 34% of output tokens in the covered accounts by Sept. 30, while ChatGPT Work accounted for 21%. The remaining 45% came from standard ChatGPT use, according to the report.

OpenAI also reported that weekly active users of Codex or ChatGPT Work among businesses with 200 or fewer users rose 8.1 times between the first week of May and the week ending Oct. 5. The report says every EU member state grew at least fivefold. Belgium recorded 12.1 times growth, Italy 10.8 times and Luxembourg 10 times. OpenAI did not provide absolute user counts alongside the growth multiple.

The reported agentic share varies across countries, from 85% in Lithuania and 81% in Cyprus to 22% in Slovenia and 26% in Austria. Germany’s share was 51%, which placed it 17th among the 27 member states. OpenAI’s figures describe activity in its own business accounts, not a representative measure of all companies or AI users in each country.

What the Token Share Shows

The report offers a view into how companies are using OpenAI’s business products, and it suggests that agent-based tools account for a substantial portion of the text those products generate. That matters as businesses weigh whether to adopt agents for coding, document work and other processes. But token share cannot establish how common agent use is across European firms, whether agents complete most tasks, or whether they improve productivity.

OpenAI’s Enterprise Signals page describes tokens as an imperfect measure of business value and notes that agentic tasks can require substantially more computation. An agent may read files, plan, produce an answer and revise it, generating far more text than a short chat response. A large token share can therefore reflect the cost and length of agent runs as well as their uptake. Readers should not treat the 55% figure as meaning that agents perform 55% of work or deliver 55% of AI’s value.

The distinction matters for public debate and business decisions. A measure based on output tokens can help track where model activity is occurring, but claims about productivity or economy-wide adoption need other evidence, such as task counts, user counts, time saved or independently measured outcomes. Those measures are not supplied with the headline figure.

A Product Launch Shapes Growth

The report’s figures cover ChatGPT Enterprise and Business accounts in the EU, and its definition of agentic use includes products introduced during the measurement period. ChatGPT Work launched in July 2026, after the small-business user comparison began in the first week of May. OpenAI says a major part of the growth followed that launch.

The timing is relevant to both headline metrics. OpenAI’s chart shows Codex rising from roughly a fifth of EU output tokens in May to 34% by Sept. 30, while ChatGPT Work contributed 21 percentage points after its launch. This does not invalidate the reported growth, but it means the figures capture adoption of a newly available product as well as any broader shift in how companies use AI.

OpenAI also compares high-use firms with typical firms. It defines “frontier firms” as the top 10% of companies by output tokens per active user. In September, EU frontier firms produced about four times as many tokens per user as typical EU firms. Frontier firms outside the EU produced more than twice as many as EU frontier firms. These are OpenAI’s customer-account comparisons, not a ranking of all companies worldwide.

What the Data Leaves Open

The report does not provide absolute user numbers for the small-business growth rate or account and user sample sizes for each country. Without those details, it is difficult to tell how much the national results may be affected by a small number of large customers, particularly in countries with smaller populations. The country percentages should be read as results for OpenAI’s observed accounts, not as a ranking of national AI readiness.

It is also unclear how many interactions or work tasks produced the 55% token share, how much value those tasks created, and how representative the customers are of EU businesses overall. The report’s figures do not resolve whether increased agent activity reflects repeated use by a limited group, wider adoption among firms, or both. No independent validation or outcome measure is included in the supplied material.

Evidence Needed Beyond Tokens

OpenAI’s report calls for European companies to move beyond early experiments by developing worker skills and confidence and making advanced capabilities more accessible and affordable. The immediate next step for readers seeking to judge that progress is to look for measures that complement token totals: participating firms and users, task-level usage, adoption over a longer period, and reported business outcomes.

Further reporting will also be needed to determine whether growth continues after ChatGPT Work’s launch effect is separated from longer-term adoption. OpenAI has not supplied those additional measures in the material described here, so whether the reported increases represent a lasting shift across European businesses remains unsettled.

Key Questions

What does OpenAI’s 55% figure measure?

It is the share of output tokens generated by Codex and ChatGPT Work in the ChatGPT Enterprise and Business accounts covered by the report, as of Sept. 30, 2026. It is not the share of users, tasks or business value.

How did OpenAI define agentic use?

The report counted activity through Codex, OpenAI’s coding agent, and ChatGPT Work, an agent mode launched in July 2026 that can work with files, browsers and company tools.

What does the 8.1-times growth figure refer to?

It refers to growth in weekly active users of Codex or ChatGPT Work among businesses with 200 or fewer users, from the first week of May to the week ending Oct. 5, 2026. OpenAI did not give the absolute number of users in the supplied report details.

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

No. The figures cover OpenAI’s EU business accounts, not all European companies. The token share also does not show what proportion of firms, workers or tasks use agents.

Source: ThorstenMeyerAI.com