OpenAI’s EU Business AI Claim Sounds Big. Check The Denominator

🔍 Read the full analysis: OpenAI’s EU Business AI Claim Sounds Big. Check The Denominator on ThorstenMeyerAI.com

Buying for a business?Offer from Amazon

Get business pricing on office and shipping supplies

• 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 coding agent Codex and ChatGPT Work generated 55% of output tokens in the company’s ChatGPT Enterprise and Business accounts across the 27 EU member states by Sept. 30, 2026. The finding, in a report published Oct. 14, measures the share of text generated by OpenAI’s models—not the share of business tasks, users or value produced with agents.

OpenAI’s Economic Research report, Europe’s Agentic Opportunity: Moving from Breadth to Depth with AI Agents, defines “agentic” use as activity from Codex and ChatGPT Work. The latter is an agent mode launched in July that can work with files, browsers and company tools. By Sept. 30, Codex accounted for 34% of output tokens and ChatGPT Work for 21%; standard ChatGPT accounted for the remaining 45%.

The company also reported rapid growth among accounts associated with businesses with 200 or fewer users. Weekly active users of Codex or ChatGPT Work in that group rose 8.1 times from the week of May 1–7 to the week ending Oct. 5. OpenAI said every EU member state recorded growth of at least fivefold. Belgium, Italy and Luxembourg had the largest reported increases, at 12.1, 10.8 and 10.0 times, respectively. The report does not give the underlying absolute user totals.

OpenAI’s country figures show substantial variation in the agentic share of output tokens: 85% in Lithuania and 81% in Cyprus, compared with 22% in Slovenia and 26% in Austria. Germany’s share was 51%, placing it 17th among the 27 countries. Those are measurements of activity within OpenAI’s accounts, not estimates of all business AI use in each country.

What the Token Share Can Show

The report offers a window into how customers of one major AI provider are using its workplace products. The figures suggest that agents are taking a growing share of OpenAI-generated output in those business accounts, and that use of Codex and ChatGPT Work expanded quickly during the period measured. For companies tracking adoption, that can be a useful signal about demand for tools that handle multi-step work.

But tokens are not a direct measure of work completed. OpenAI’s Enterprise Signals page calls tokens “an imperfect measure of business value” and says agentic tasks can require substantially more computation. An agent may inspect files, plan, draft code or documents, check results and retry. That process can generate far more text than a short chat response, even if it represents only one task. The 55% figure therefore cannot establish that most business interactions—or most value—now come from agents.

The distinction matters for readers comparing adoption claims. A measure based on output volume can reflect longer or more iterative model runs, as well as changes in the products available. It does not by itself show that companies are more productive, that workers have shifted most of their work to agents, or that European businesses broadly have reached the same level of adoption.

Popular right now

De'Longhi Rivelia Fully Automatic Espresso Machine with Milk Frother, White

AmazonView Latest Price

Ninja Flip Toaster Oven & Air Fryer, 8-in-1, Stainless Steel, SP151

AmazonView Latest Price

Samsung 65-Inch The Frame Pro Neo QLED 4K TV (2026) Bundle with Modern Brown Bezel, HDMI Cables, and Home Theater Guide

AmazonView Latest Price

As an affiliate, we earn on qualifying purchases.

A July Launch Shapes the Trend

The period behind the headline includes the launch of ChatGPT Work in July 2026. OpenAI’s chart, as described in the source report, shows Codex accounting for roughly a fifth of EU output tokens in May and 34% by Sept. 30. ChatGPT Work then contributed 21 percentage points by the end of September. The product’s arrival accounts for more than half of the roughly 35-point rise in the combined agentic share described in the analysis.

The same timing affects the reported small-business growth. The comparison begins with the first week of May, before ChatGPT Work launched, and ends in early October. OpenAI says a major part of the growth followed the introduction of ChatGPT Work. The 8.1-times figure is a ratio between weekly active-user counts at two points in time; without the underlying counts, readers cannot tell how large the starting base was or how many businesses the increase represents.

OpenAI also compares its highest-use customers with other groups. It defines “frontier firms” 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. OpenAI says EU companies make up about 10% of frontier firms worldwide, compared with about 65% in North America. These are comparisons within OpenAI’s customer data, not a census of companies across regions.

The Missing Counts Behind the Charts

The report does not provide absolute user or account counts for the small-business growth calculation, or sample sizes for each country. That limits what readers can infer from the 8.1-times increase and the country rankings: a large multiple can start from a small base, while a national percentage can be shaped by a limited number of large customers. OpenAI itself notes that smaller countries including Cyprus, Estonia and Lithuania have relatively high agentic use within their larger firms.

The country charts also do not establish national readiness or rank all companies in the EU. The available figures describe OpenAI accounts, and the source material provides no independent data with which to test how representative those accounts are. It is also unclear how much of the change reflects new customers, increased use by existing customers, product availability, or a combination of those factors.

Finally, the report does not show whether greater token use led to measurable productivity gains, lower costs or improved business outcomes. The source analysis offers a hypothetical illustration that assumes agent runs produce 10 to 30 times the output of a chat reply; that is the author’s assumption, not a measurement from OpenAI’s report. The illustration should not be treated as an estimate of the actual share of agentic interactions.

Evidence Needed Beyond Token Growth

OpenAI’s report calls for businesses to build workers’ skills and confidence, move beyond experimentation and make advanced capabilities more accessible and affordable. The company’s next public updates will be more informative if they provide absolute account and user counts, explain how the customer mix changes over time, and distinguish new adoption from increased activity among existing customers.

For a clearer view of business impact, future evidence would also need to connect product use to task completion, time saved, costs or other outcome measures. Until such data are available, the current findings support a narrower conclusion: agent products generated a majority of output tokens in the OpenAI EU business accounts covered by the report, while the broader effects on European businesses remain unestablished.

Key Questions

What does OpenAI’s 55% figure measure?

It is the share of output tokens generated by Codex and ChatGPT Work in OpenAI’s ChatGPT Enterprise and Business accounts across the EU by Sept. 30, 2026. It is not the share of users, tasks or business value.

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

No. The data cover OpenAI’s EU business accounts, not every business in the EU. The report does not provide a representative estimate of agent use across all European companies.

How did small-business agent use change?

OpenAI reported that weekly active users of Codex or ChatGPT Work among businesses with 200 or fewer users increased 8.1 times between the first week of May and the week ending Oct. 5, 2026. It did not publish the absolute user counts.

Why does the July launch matter?

ChatGPT Work launched in July and accounted for 21 percentage points of EU output tokens by Sept. 30, according to the report. That means the product’s availability contributed substantially to the measured rise in agentic output.

Does higher token use prove higher productivity?

No. The report does not establish productivity gains or other business outcomes. OpenAI describes tokens as an imperfect measure of business value, and agent tasks can generate more text because they involve multiple steps.

Source: ThorstenMeyerAI.com