A Real AI Workforce Rollout: What Univé Did — and What Small Teams Can Take From It

What happened

OpenAI published a case study on how Univé, a Dutch insurance cooperative, built what it calls an "AI-ready workforce" using ChatGPT Enterprise. According to the case study, Univé's approach combined leadership commitment with a governance framework and leaned on employee-led discovery rather than assigning AI tasks from above. Staff were encouraged to identify their own use cases within structured guidelines rather than receiving a prescribed list of approved applications.

The case study is written by OpenAI and should be read with that in mind. That said, the structural lessons — starting with willing early adopters, establishing data and usage guidelines before broad rollout, and measuring actual time savings rather than assuming them — reflect patterns that show up consistently in organizations where AI adoption produces real results rather than nominal participation.

What it means for your business

You do not need an enterprise contract to run a version of this approach. Identify one or two employees who are already curious about AI tools and give them latitude to experiment within a defined boundary: specific tasks, no sensitive customer data sent to external tools, outputs reviewed before going to clients. Set basic written guidelines early — even a one-page document covering what stays out of AI tools reduces risk and ambiguity. Expect uneven uptake; some staff will find clear uses immediately, others won't, and forcing adoption rarely produces genuine use. The goal is not to make everyone an AI power user — it is to find the handful of workflows where these tools genuinely reduce time or effort, and build from there.

Source: OpenAI