OpenAI says agentic AI usage is rising quickly across the EU, based on its latest economic research. The company reports that among ChatGPT Enterprise customers in its EU sample, agentic use already represents a substantial share of activity, and that Codex and ChatGPT Work together account for more than half of output tokens.
This matters because it offers a concrete view into how organizations are actually using AI: less as a question-answer tool and more as an operational system that can plan and carry out work. That shift has implications for software teams, procurement, and governance.
The practical takeaway is that if you build AI products for enterprises, you should think in terms of workflows, output volume, and controls. The market signal is clear: agents are becoming part of daily work, not a side feature.
Source: OpenAI report