Understanding The Scope Of OpenAI’s EU Agentic AI Claim

🔍 Read the full analysis: Understanding The Scope Of OpenAI’s EU Agentic AI Claim 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 said Codex and ChatGPT Work generated 55% of output tokens in its ChatGPT Enterprise and Business accounts across the European Union by September 30, 2026, a figure the company uses to describe the growth of workplace AI agents. The report covers OpenAI’s own customers, and token share does not establish how many business tasks use agents or how much value they produce.

OpenAI’s Global Affairs team published The EU’s next chapter in AI is agentic on October 14, alongside a ten-page report from OpenAI Economic Research, Europe’s Agentic Opportunity: Moving from breadth to depth with AI agents. The analysis covers ChatGPT Enterprise and ChatGPT Business accounts in the 27 EU member states. OpenAI defines agentic use in this report as use of Codex, its 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%, according to the report; ordinary ChatGPT accounted for the remaining 45%. OpenAI also reported that weekly active users of Codex or ChatGPT Work at businesses with 200 or fewer users grew 8.1 times between the first week of May and the week ending October 5. The report says every EU member state recorded at least fivefold growth. It does not provide absolute user counts.

The agentic share varied by country. OpenAI reported 85% in Lithuania and 81% in Cyprus, compared with 22% in Slovenia and 26% in Austria. Germany’s share was 51%, ranking 17th among the 27 countries. These figures describe usage within OpenAI’s accounts, not the full population of businesses in each country.

What the Token Share Shows

The report gives businesses and policymakers a view into how some companies are using OpenAI’s workplace products, with a notable shift toward tools that can carry out multi-step work. But the headline number has a narrower meaning than “most AI use is now agentic”: it is a share of generated output tokens, not a share of interactions, users, tasks or business outcomes.

OpenAI’s Enterprise Signals page calls tokens an imperfect measure of business value and notes that agentic tasks can require substantially more computation. An agent may read files, plan, draft code or documents, check its work and retry, generating much more text than a short chat response. That difference can raise agents’ share of tokens without a corresponding increase in their share of tasks.

The figures also matter because the report’s proposed next step is not merely wider access. OpenAI argues that Europe should turn broad adoption into deeper use by building worker skills and confidence and making advanced tools more accessible and affordable. The data documents activity among OpenAI customers; it does not independently show that deeper use would raise productivity or improve business results.

How OpenAI Defines Agent Use

The reported 55% combines two products with different timelines. Codex’s share of EU output tokens rose from roughly a fifth in May to 34% by September, according to the report’s chart. ChatGPT Work launched in July and accounted for 21% by September 30. The report therefore captures both growing use of Codex and a new product entering the mix during the measurement period.

That timing also shapes the small-business growth figure: its baseline is the first week of May, before ChatGPT Work launched. OpenAI says a major part of the growth followed the launch. The 8.1-times figure is a comparison with that starting week, not an absolute count of new users, and the report does not separate how much growth came from product availability versus other factors.

OpenAI also compares high-use “frontier firms,” defined 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. The report estimates EU companies make up about 10% of frontier firms worldwide, while North American companies account for about 65%. These comparisons use OpenAI’s token-based definition of activity.

What the Data Cannot Establish

The figures do not establish what proportion of business tasks is agentic, whether users complete work faster, or whether the tools create measurable commercial value. OpenAI itself warns that tokens are an imperfect proxy and says agentic tasks can consume more computation. Any estimate of agent interactions from the 55% token share would require assumptions about how much output an agent run produces relative to a chat exchange; the report does not provide that comparison.

Country-level interpretation is also limited. OpenAI does not disclose the number of accounts or users behind each country figure, so the influence of a small number of large customers is unknown. The rankings should not be treated as a national readiness league table. The report also does not give absolute user numbers for the small-business growth statistic or identify how much of the increase reflects the July launch.

Finally, the evidence comes from a company describing use of its own products. It provides a detailed look at OpenAI’s customer base, but the supplied figures do not establish how representative those customers are of all EU businesses or how the results compare with other providers.

From Adoption to Deeper Use

OpenAI’s report calls for businesses to build workers’ skills and confidence and make advanced AI capabilities more accessible and affordable. It does not set out a timetable or a measured target for achieving those aims. The next useful evidence would include follow-up data showing whether the user growth persists after the initial launch period and whether companies report changes in task completion, costs or output.

For readers tracking the claim, the key comparison will be whether later reports add absolute user counts, disclose more about country-level samples and distinguish tokens from completed tasks and business results. Until then, the figures support a narrower conclusion: agent tools are generating a growing share of output in OpenAI’s EU business accounts, while their broader economic impact remains unmeasured in this report.

Key Questions

What does OpenAI mean by a 55% agentic share?

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

Which tools count as agentic in the report?

OpenAI counts Codex, its coding agent, and ChatGPT Work, an agent mode launched in July 2026 that can work with files, browsers and company tools.

How quickly did smaller businesses adopt the tools?

OpenAI says weekly active users of Codex or ChatGPT Work at businesses with 200 or fewer users rose 8.1 times from the first week of May to the week ending October 5. The report does not provide absolute user counts, and the period includes ChatGPT Work’s launch.

Does the report show that agents improved productivity?

No. It reports product usage and output tokens, not measured productivity or business value. OpenAI describes tokens as an imperfect measure of business value.

Do the country figures represent all businesses in the EU?

No. They cover OpenAI’s ChatGPT Enterprise and Business accounts. OpenAI does not provide country-level account or user counts in the supplied report, so the figures should not be read as representative measures of every country’s businesses.

Source: ThorstenMeyerAI.com