Staff need room to ask questions, practise relevant tasks and learn how to check an AI output. Training should make the expected workflow clear and give people support after the session.
Start with responsibilities and concerns
Ask what each role does, which tasks feel repetitive and what people need to trust a new process. Explain the intended use, the limits and who remains accountable. Avoid judging confidence by job title or age; assess what each person needs to learn.
Use role-specific practice
A marketing exercise might start with an approved brief. A service exercise might use a synthetic enquiry. An operations exercise might prepare a checklist. Use samples that let staff identify errors and refine the draft without exposing confidential information.
Teach review alongside prompting
- Define the task, audience, constraints and output format.
- Check statements against approved source material.
- Identify missing context and unsupported claims.
- Use the agreed human approval and escalation process.
- Record helpful changes in a shared resource.
Keep adoption measurable
Provide reusable prompts, examples and operating notes. Agree who maintains them and how staff raise questions. Review task quality, time including correction, and whether the process is being used. Set targets from a measured baseline rather than assuming a universal improvement percentage.
Learn about workshops and team training or use the Readiness Assessment to identify an initial gap.

