The Key Question Behind OpenAI’s Agentic AI Claim For EU Businesses

Read the full analysis: The Key Question Behind OpenAI’s Agentic AI Claim For EU Businesses on ThorstenMeyerAI.com

OpenAI says its AI agents generated 55% of output tokens in its ChatGPT Enterprise and Business accounts across the European Union by September 30, a measure of growing use that does not show how many tasks agents handle or what business value they produce. The company’s October 14 report also says small-business use of its agent products rose sharply after a new product launched, while its most intensive EU business users still trail the global frontier.

OpenAI Economic Research’s report, Europe’s Agentic Opportunity: Moving from breadth to depth with AI agents, covers accounts in the 27 EU member states. It defines agentic use as activity from Codex, OpenAI’s coding agent, and ChatGPT Work, an agent mode launched in July 2026 that can work with files, browsers and company tools. By September 30, Codex accounted for 34% of output tokens and ChatGPT Work for 21%; ordinary ChatGPT accounted for the remaining 45%.

Among businesses with 200 or fewer users, weekly active users of Codex or ChatGPT Work increased 8.1 times from the first week of May to the week ending October 5, according to OpenAI. The company reports at least fivefold growth in every member state, with Belgium at 12.1 times, Italy at 10.8 times and Luxembourg at 10 times. The report does not provide absolute user totals, and says a major part of the increase followed ChatGPT Work’s launch.

OpenAI also compares its highest-use customers, defined as the top 10% by output tokens per active user. In September, EU frontier firms generated about four times as many tokens per user as typical EU firms. Frontier firms outside the EU generated more than twice as many as the EU frontier group. OpenAI estimates that EU companies make up about 10% of frontier firms worldwide, compared with about 65% in North America.

What the Token Share Can Show

The figures offer a view of how quickly customers of one major AI provider are taking up its agent products. That matters to EU businesses weighing investment in software, staff training and workplace processes. OpenAI’s data suggests that usage is expanding, especially among smaller customers, but it also points to a gap between ordinary business use and the most intensive users globally.

The 55% figure is a token share: it counts the text generated by OpenAI’s models, not the percentage of employees using agents, the number of work tasks completed, or measured savings and revenue. Agent runs can involve planning, reading files, producing drafts and checking results, which may generate more tokens than a short chat response. OpenAI itself calls tokens an imperfect measure of business value and says agentic tasks can require substantially more computation.

That distinction limits what the headline can establish. The report supports a claim about where model-generated output is going inside OpenAI’s customer accounts. It does not establish that most EU business AI work is agent-driven, that agents have improved productivity, or that the reported growth represents a lasting change across the region’s economy.

A New Product Shapes the Trend

The period behind the growth figures includes a product launch. According to OpenAI’s chart, Codex accounted for roughly a fifth of EU output tokens in May and 34% by September 30. ChatGPT Work was introduced in July and reached 21% by the end of September. The report’s own account says much of the small-business increase came after that launch.

That timing matters when interpreting growth from the first week of May. An 8.1-fold rise measures activity against a period before ChatGPT Work was available, so it combines adoption over time with the arrival of a new tool. OpenAI provides no absolute small-business user counts in the cited material, making it difficult to assess the size of the starting base or the number of businesses represented.

Reported agentic output also varies substantially by country: Lithuania was at 85% and Cyprus at 81%, while Slovenia was at 22% and Austria at 26%. Germany was at 51%, ranking 17th among the 27 member states. OpenAI does not publish country-level account or user counts in the material provided, so these figures describe its customer base and should not be read as a definitive ranking of national AI adoption.

What OpenAI’s Data Leaves Out

The report does not disclose the absolute number of accounts or users behind its country comparisons, or the total number of small-business users behind the 8.1-fold increase. Without those figures, readers cannot judge how representative the data is or how much growth came from a small initial group. The source material also does not provide independent verification of the account-level results.

There is no direct measure here of completed tasks, time saved, accuracy, costs or financial returns. The token measure may rise because agents perform longer, more computationally demanding runs, rather than because they handle a larger share of business activity. It is also unclear whether usage will persist after the launch period or translate into wider adoption beyond OpenAI’s paying business accounts.

The report’s country figures have an additional limitation: customer mix may differ from one market to another, and OpenAI provides no sample sizes to show how much a few large accounts affect national percentages. The reported gap between EU and non-EU frontier firms is likewise a comparison of OpenAI-defined usage groups, not a broad measure of company productivity or the entire EU economy.

From Early Use to Measured Results

OpenAI’s report calls for businesses to build workers’ skills and confidence and make advanced AI capabilities more accessible and affordable. The company’s next useful evidence would include follow-up data showing whether small-business use continues beyond the launch period, alongside absolute user counts and clearer descriptions of the customer samples in each country.

For businesses and policymakers, the key test is not only whether agents generate more output, but whether they reliably complete valuable work at an acceptable cost. The October report does not answer that question. Further measurement of task outcomes, productivity, errors and spending would help distinguish fast-growing product use from sustained economic impact.

Key Questions

What does OpenAI mean by a 55% agentic share?

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

How did OpenAI define agentic use?

The report counts activity from Codex, OpenAI’s coding agent, and ChatGPT Work, its agent mode launched in July 2026, which can work with files, browsers and company tools.

Did small-business agent use grow 8.1 times?

OpenAI reports 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 October 5. The company does not give absolute user totals, and says much of the growth followed ChatGPT Work’s launch.

Does the report show that agents improved productivity?

No. The report describes product use and generated tokens, but the cited material does not measure completed tasks, time saved, accuracy, costs or financial returns.

Source: ThorstenMeyerAI.com