What To Check Before Believing OpenAI’s Agentic AI Figure For EU Business

🔍 Read the full analysis: What To Check Before Believing OpenAI’s Agentic AI Figure For EU Business on ThorstenMeyerAI.com

Before you orderOffer from Amazon

Get the latest gadgets delivered free with Prime

• Fast, free delivery on millions of items
• Prime Video, Amazon Music and more included
• Member-only deals all year

Start your free Prime trial →Free trial for eligible customers · Cancel anytime

As an affiliate, we earn on qualifying purchases.Hide

OpenAI says its agents generated 55% of output tokens in the company’s ChatGPT Enterprise and ChatGPT Business accounts across the European Union by September 30, 2026. The figure, drawn from OpenAI customers rather than the wider European economy, measures text generated by the models—not the share of users, work tasks or business value handled by agents.

OpenAI Economic Research published Europe’s Agentic Opportunity: Moving from Breadth to Depth with AI Agents on October 14, alongside a post from the company’s Global Affairs team. The report defines agentic use as activity by Codex, OpenAI’s coding agent, and ChatGPT Work, an agent mode launched in July that can work with files, browsers and company tools. By September 30, Codex accounted for 34% of output tokens in the covered accounts and ChatGPT Work for 21%; standard ChatGPT accounted for the remaining 45%.

For businesses with 200 or fewer users, OpenAI reports that weekly active users of Codex or ChatGPT Work increased 8.1 times between the first week of May and the week ending October 5. The report says every EU member state grew at least fivefold over that period. It does not provide absolute user counts, and it notes that a major part of the increase followed ChatGPT Work’s launch.

OpenAI also reports wide differences among countries. Agentic output-token shares ranged from 85% in Lithuania and 81% in Cyprus to 22% in Slovenia and 26% in Austria. Germany’s share was 51%, placing it 17th among the 27 member states. These are measurements of OpenAI’s own business-account activity, not country-level estimates of all business AI use.

What Token Share Can Tell Businesses

The report offers a rare, detailed view of how one major AI provider’s business customers use its products. Its figures indicate that agents account for a substantial share of OpenAI’s generated output in EU business accounts, and that activity in smaller accounts rose quickly during a period when the company introduced a new agent product.

But output tokens are a limited proxy for workplace adoption. Agent runs can involve reading files, planning, generating material and checking results, producing far more output than a short chat exchange. OpenAI’s Enterprise Signals page calls tokens “an imperfect measure of business value” and says agentic tasks can require substantially more computation. The 55% figure therefore cannot establish that most AI interactions or business tasks are agent-led, or that agents have delivered productivity gains.

That distinction matters to business leaders weighing investment and to policymakers considering Europe’s competitive position. A rise in usage may show that customers are trying or relying on a product; it does not, by itself, show improved output, lower costs or economy-wide adoption. OpenAI’s policy argument—that Europe should turn broad adoption into deeper use—rests on evidence about its customer activity, not an independent survey of European firms.

How OpenAI Defines Agent Use

The report covers ChatGPT Enterprise and ChatGPT Business accounts in the EU’s 27 member states. Its central comparison is based on output tokens, the text generated by the models. It counts Codex and ChatGPT Work as agentic tools and compares their output with standard ChatGPT. The report’s date range matters: the small-business growth comparison begins in early May, before ChatGPT Work launched in July, and ends in early October.

OpenAI also compares high-usage companies with typical users. It defines “frontier firms” as the top 10% 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 the EU frontier. OpenAI says EU firms make up about 10% of frontier firms worldwide, compared with about 65% in North America. Those figures describe the company’s account activity under its own definitions, rather than a census of firms or a direct measure of innovation.

What the Report Cannot Establish

The report does not disclose the number of accounts or users behind each country’s results, making it difficult to judge how representative the national rankings are. OpenAI itself notes that some smaller countries have relatively high agentic use within larger firms. Without country-level sample sizes, small numbers of large customers could affect the shares.

The publication also gives no absolute count of small-business agent users, so the 8.1-times increase cannot show how many firms adopted the tools or how broad that adoption is. The comparison begins before ChatGPT Work was available, and the report attributes a major part of growth to its introduction. It remains unclear how much use will persist beyond the initial adoption period.

Nor does the token data measure business outcomes. The report does not establish whether agent use saved time, reduced costs or improved work quality. It also does not represent companies that do not use OpenAI products, so its figures cannot independently settle how far European businesses as a whole have adopted agents or whether EU firms are falling behind globally.

Evidence Needed Beyond Usage

The report sets out an argument for expanding worker skills, confidence and access to advanced tools, but it does not provide a timetable or measurable targets for those steps. The next useful evidence would include absolute user and account counts, clearer sampling information by country, and repeated measurements that show whether activity continues after product launches.

For businesses and policymakers, a fuller assessment would also require evidence tied to outcomes: which tasks agents handle, how often people review their work, and whether organizations see measurable changes in time, cost or quality. Until such data are available, OpenAI’s findings are best read as a snapshot of agent activity among its EU business customers—not a measure of agent adoption or economic impact across Europe.

Key Questions

What does OpenAI mean by 55% of output being agentic?

It means Codex and ChatGPT Work generated 55% of output tokens in the EU ChatGPT Enterprise and Business accounts covered by the report by September 30, 2026. It does not mean agents handled 55% of users’ tasks or created 55% of business value.

How did small-business agent use change?

OpenAI reports an 8.1-times increase 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 October 5, 2026. The report provides no absolute user count and says a major part of the growth followed ChatGPT Work’s launch.

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

No. It covers OpenAI’s ChatGPT Enterprise and Business accounts in the EU, not all European businesses. The token figures also do not show what proportion of firms, tasks or workers use agents.

Can the country rankings show which EU nations are most advanced in AI?

Not on their own. OpenAI does not give account or user sample sizes for each country, and its rankings reflect the company’s customer mix. The results should not be treated as a definitive league table of national AI readiness.

Source: ThorstenMeyerAI.com