Understand the current literacy provision.
Amended Article 4 requires providers and deployers to support the development of AI literacy for relevant staff and other people acting on their behalf, taking knowledge, experience and context into account. It does not require guaranteeing a specific literacy level for every individual.
Use that context to design a practical programme rather than assuming that one certificate or identical course is the only acceptable approach.
Start with the decisions each role makes.
A general user needs to recognise approved tools, suitable inputs and the need to check outputs. A procurement reviewer needs to ask useful supplier questions. An oversight owner needs to understand intervention, escalation and the limits of the system being reviewed.
Map those needs before selecting materials. Include contractors or other relevant people where they operate or use systems on the organisation's behalf.
Use realistic, safe examples.
Discuss situations that staff actually encounter: an uncertain answer, a request to paste confidential information, a new AI feature or an output influencing a consequential decision. Use hypothetical or appropriately prepared material, and show where to ask for help.
- Approved uses and prohibited or restricted activities.
- Known limitations and output-review expectations.
- Handling of sensitive information.
- Routes for declaring new uses and reporting concerns.
Record support and improve it over time.
Keep the programme scope, materials, attendance or participation information where appropriate and feedback about confusing tasks. Look for practical gaps rather than treating a completed training field as proof that every person can handle every scenario.
EUAIC can help connect learning records to roles and use cases within an agreed governance scope. Update the material when tools, policies or responsibilities change, and keep competence decisions for specialist roles with the appropriate accountable managers.