OpenAI report: agentic tools are now over half of EU enterprise AI use

OpenAI Economic Research published a report on October 14 examining how agentic use is spreading across European enterprises and small businesses, focusing on differences in usage intensity and the kinds of work people bring to ChatGPT. The framing is blunt: EU businesses are moving beyond asking questions to using agents that carry tasks from planning through completion.

The strongest figures come from enterprise customers. Among the EU ChatGPT Enterprise accounts in the sample, Codex and ChatGPT Work now account for more than half of output tokens. The analysis defines small and medium businesses as those with 200 or fewer associated users.

Growth in the smaller-business segment has been steepest. Since May 2026, weekly active users of Codex and ChatGPT Work among EU small and medium businesses grew more than eightfold, with growth recorded in every member state.

Adoption is far from uniform, however. Within the enterprise sample, the share of output tokens produced through Codex and ChatGPT Work ranged from 22% in Slovenia to 85% in Lithuania, and the kinds of work people bring to the tools vary by country as well.

The report defines frontier firms as the top 10% of companies by output tokens per active user. In September, frontier firms in the EU generated more than four times as many output tokens per active user as typical EU firms, yet frontier firms outside the EU generated more than twice as many per active user as EU frontier firms. EU-based companies made up only 10% of global frontier firms, even though the EU accounts for a higher share of registered companies worldwide.

OpenAI chief financial officer Sarah Friar said in the report that Europe is now in a new era of AI in which tools can carry out tasks for people, that more than half of AI use among the company's EU enterprise and business customers involves such tools, and that small business users have grown eightfold across the region even as plenty of room for growth remains.

The report pulls the analysis toward depth rather than breadth. The gap it identifies is not about adoption but about intensity, and closing it means helping businesses move past experimentation, giving workers the skills and confidence to use AI well, and making advanced capabilities more accessible and affordable.

What to watch next is whether those numbers hold as agentic products spread. Because the sample covers only ChatGPT Enterprise and Business accounts, the 22%-to-85% spread across member states is the metric to track when judging whether Europe's AI productivity story is actually narrowing.