In forceAI literacy8 min read

AI Literacy Under Article 4: A Practical Employer Checklist

As amended in 2026, Article 4 requires AI providers and deployers to take measures that support the development of AI literacy among staff and other people dealing with AI systems on their behalf. It does not require them to guarantee a specific level. Measures should reflect the relevant people's knowledge, experience, education and training, the context of use and the people or groups affected.

Applies from: 2 February 2025
For: Employers, AI deployers, AI providers, HR and compliance teams
Last legally reviewed: 30 September 2026

Educational guidance: This overview explains EU rules in general terms. Applicability depends on the organisation, system, service and national law.

What Article 4 actually asks organisations to do

Article 4 requires providers and deployers to take measures that support the development of AI literacy among the relevant people. The amended text expressly says they are not required to guarantee a specific level of AI literacy. Measures should take account of technical knowledge, experience, education and training, the context of use and the people or groups on whom the systems are used.

The appropriate programme therefore depends on the people, tools and risks involved. A marketing employee using a writing assistant, an HR team using ranking software and an engineer integrating a model API do not need identical training.

Who should be included

The scope is wider than permanent employees. It can include contractors, temporary staff and other people operating or using AI systems on the organisation's behalf.

People who select, buy or approve AI systems.
People who enter information or instructions into AI tools.
People who review, rely on or communicate AI outputs.
Managers responsible for AI-supported processes.
Technical teams that integrate, monitor or change AI systems.
External workers dealing with AI for the organisation.

Build training around real work

A useful programme starts with approved tools and actual tasks. It should explain the limits of outputs, confidential and personal data, human review, bias and discrimination, transparency, escalation and the consequences of using an AI system outside its approved purpose.

Higher-impact uses need deeper role-specific material and exercises. Training should be refreshed when the system, task, legal requirements or affected group changes materially.

Evidence without inventing a certification rule

The Regulation does not set a universal number of training hours or require a particular certificate. Organisations should still retain enough evidence to show that their measures were deliberate and proportionate.

A practical evidence file can record who received the measure, the training date, content and source version, the person's role and relevant AI risks, attendance or completion, the responsible owner, the review date and the triggers for refresher training. These records are useful evidence; the Act does not prescribe one mandatory certificate or record format.

Action list

Practical checklist

List the people who deal with each AI system.
Assess their knowledge, role, context and affected groups.
Define role-specific learning outcomes.
Teach approved use, limitations, data handling and human review.
Record participants, dates, content versions, attendance or completion and the responsible owner.
Set review dates and triggers for refresher training and reassessment.

Questions

Common questions

Does Article 4 require every employee to complete the same course?

No. Measures should be proportionate to the person's knowledge, experience, education, training and the context in which the AI system is used.

Do contractors need AI literacy training?

They may fall within scope when they deal with AI systems on the organisation's behalf. The organisation should assess their role rather than limiting the programme to payroll employees.

How often must AI literacy training be repeated?

Article 4 does not impose one universal refresh interval. Organisations should review training when tools, use cases, risks, users or applicable guidance change.

Primary material

Official sources

Last legally reviewed: 30 September 2026. Recheck the consolidated legislation and relevant national implementation before relying on this guide for a specific decision.