OpenAI’s latest economic research paints a clear picture: agentic AI adoption is expanding across European businesses, shifting usage from simple question-and-answer interactions to systems that can perform tasks end-to-end. The company’s analysis of ChatGPT Enterprise and Business customers in the EU reveals that agentic tools are already central to many workflows, even as uptake varies by country and firm type.
In the EU sample examined, Codex and ChatGPT Work together now generate more than half of output tokens among ChatGPT Enterprise customers. That concentration of output indicates a substantive move toward agentic processes—where AI agents carry out actions on behalf of users—rather than limiting use to conversational queries. The prevalence of these tools suggests firms are leveraging agentic capabilities for coding, automation and other task-driven workflows.
The report highlights particularly strong growth among small and medium businesses (SMBs). OpenAI defines SMBs in the analysis as firms with 200 or fewer associated users. Since May 2026, weekly active users of Codex and ChatGPT Work among EU SMBs grew by more than eightfold, with increases reported across every EU member state included in the sample. That rapid expansion among smaller organizations suggests agentic tools are becoming accessible beyond large enterprises and are being integrated into a broader set of day-to-day operations.
Adoption is nonetheless uneven across the region. Among the ChatGPT Enterprise accounts analyzed, the share of output tokens produced through Codex and ChatGPT Work ranged widely by country—from roughly 22% in Slovenia to about 85% in Lithuania. The analysis also notes that the types of tasks organizations bring to ChatGPT differ between member states, reflecting varying sectoral needs, local practices and possibly differing maturity in digital workflows.
OpenAI’s report also compares activity within the EU to global patterns by identifying “frontier firms”—the top 10% of organizations by output tokens per active user. In September, frontier firms within the EU sample produced more than four times as many output tokens per active user as a typical EU firm, demonstrating a steep internal productivity gradient. However, EU frontier firms still trail their non-EU counterparts: frontier firms outside the EU generated more than twice as many output tokens per active user as frontier EU firms. The analysis further finds that EU-based firms make up only 10% of the global frontier group, a share OpenAI characterizes as low relative to the EU’s global share of registered companies.
OpenAI frames these findings as evidence of opportunity as much as a diagnosis of uneven adoption. The company suggests that moving from preliminary adoption to deeper, intensive use of agentic tools could strengthen firms, increase productivity and deliver benefits to workers. Closing the gap, the report argues, will involve helping businesses move beyond experimentation, equipping workers with the skills and confidence to deploy AI effectively, and making advanced tools more accessible and affordable.
The policy and organizational implications are presented as priorities rather than prescriptive measures. By documenting where agentic tools are already integrated and where uptake lags, the analysis provides a baseline for efforts aimed at expanding capability access, workforce training and practical applications across sectors and member states. That baseline is intended to inform both public policymakers and private leaders as they consider how to support broader and more effective deployment of these systems.
OpenAI’s finance chief highlighted the transition as a new chapter for European AI use, noting that more than half of AI activity among the company’s EU enterprise and business customers now involves agentic tools and that small business user numbers have multiplied across the region. She emphasized the potential for broader adoption to support stronger businesses and economic growth, provided workers are given the tools and skills needed to use these systems effectively.
As agentic AI adoption continues to accelerate in the EU, the report underscores a central question for leaders in business and government alike: how to translate expanding use into sustained, equitable gains in productivity and worker capability. The analysis offers a data-driven starting point for that conversation, highlighting both progress to date and the work still required to ensure the benefits of agentic systems are realized widely across the European economy.
Source: Read the original source
Leave a Reply