Preview
Rolling out ChatGPT in your company
How individual accounts become a governed way of working: choosing the first teams, setting rules, connecting data sources, training on real tasks and measuring the benefit. The guide is currently available in German.
What the guide covers
- 1
Starting point
Many employees have long been using ChatGPT, often with personal accounts and without rules. This chapter shows how to take an honest look at where you stand today, define the goal in one sentence, name the people responsible and choose between ChatGPT Business and Enterprise.
- 2
The first teams
The first team is both the proof and the template for all the others. You'll learn how to recognize a suitable team, how to define three use cases, each with a short profile and a metric, and why decisions about people don't belong at the start.
- 3
Rules and data protection
Which data may go in, and which may not? This chapter walks through data classes, data processing agreements under the GDPR, the obligations under the EU AI Act, the works council's co-determination rights and the workspace settings. It includes a one-page template for AI usage rules.
- 4
Data sources and connections
ChatGPT becomes useful once it knows what your company knows. This chapter describes the stages, from an uploaded file to projects and custom GPTs to connecting SharePoint and line-of-business systems, and what should be checked before each connection.
- 5
Training on real tasks
Having access is not the same as having the skills. You'll read how foundational training and working sessions on the team's own tasks are structured, what a good prompt looks like and how to document the training as evidence of AI literacy.
- 6
Measuring and expanding
After a few weeks, management decides on the next step. This chapter shows how to measure usage, benefit and acceptance without monitoring individuals, and how the next team builds on the first. The appendix contains the two-week rollout plan for the first team.