People’s starting point
Technical knowledge, experience, education, and training should shape the measures provided.
EU AI Act · Article 4
Build your AI literacy programme on evidence.
Under Article 4 as amended by Regulation (EU) 2026/1744, providers and deployers must take measures that support the development of AI literacy among staff and others operating AI systems on their behalf, considering their knowledge, experience, training, and use context. AILAT can establish a structured workforce baseline for deciding what learning should come next.
AILAT supports programme design and evaluation. It is not a legal compliance certification and does not replace organization-, role-, or system-specific training.
Legal-status note
Reviewed 18 August 2026
Article 4 has applied since 2 February 2025.
Amended by Regulation (EU) 2026/1744, in force since 27 July 2026.
The amended duty is to take measures that support the development of AI literacy; no particular level must be guaranteed for each person.
The law and official guidance are evolving. This page gives general information, not legal advice. Check the current Official Journal text and applicable national enforcement rules for your organization.
The right measures are contextual, not one-size-fits-all.
The amended Article 4 wording connects literacy measures to the people involved and the context in which AI is used. The Commission’s practical guidance similarly emphasizes the organization’s role and the risks associated with its systems.
The organization’s role
Provider and deployer responsibilities, plus how people operate or use AI on the organization’s behalf.
Systems and context
The AI systems involved, the context in which they are used, and the opportunities and risks people need to understand.
People who may be affected
The persons or groups on whom the AI systems are used should inform the depth and focus of literacy measures.
No mandatory test, certificate, course, or AI officer.
The Commission’s guidance confirms that Article 4 allows flexibility. Measuring every employee’s knowledge is not mandatory, and an assessment by itself is not evidence that every relevant duty has been met.
Assessment
An assessment can establish a baseline, but the AI Act does not require one specific test.
Certificate
The Commission says there is no need for a specific certificate; internal records of training or guidance can be relevant evidence.
Uniform curriculum
Different roles and systems may justify different learning depth, content, and delivery.
Compliance shortcut
Course completion or a literacy score cannot replace system, risk, policy, and governance work.
Move from context to evidence, then from evidence to action.
This sequence is a planning model, not a legal checklist. It makes the baseline useful by placing it inside the organization’s own AI context.
- 01
Map the context
Identify the AI systems in use, who works with them, the organization’s role, and the people affected.
- 02
Establish a baseline
Measure relevant knowledge and reasoning across the workforce or a defined cohort.
- 03
Define what each group needs
Combine baseline findings with role-, system-, and risk-specific requirements.
- 04
Deliver suitable measures
Provide learning, guidance, policies, and practical support appropriate to each target group.
- 05
Review and keep records
Reassess where useful and retain internal evidence of the measures taken and why.
A workforce baseline is one evidence input—not the whole programme.
AILAT can reveal current knowledge patterns and learning priorities while leaving system classification, policy, governance, and role-specific instruction with the organization and its advisers.
AILAT can help you
- Establish a 20-question baseline across four AI literacy dimensions
- Give each participant a private learning-oriented report
- See privacy-protected organization, role, or department patterns
- Prioritize shared learning needs and retain an aggregate report
AILAT does not
- Determine whether you are legally a provider or deployer
- Classify your AI systems or evaluate their controls
- Replace system-specific instructions or oversight training
- Certify Article 4 compliance
Article 4 AI literacy FAQ
Does Article 4 require an AI literacy assessment?
No. The AI Act does not prescribe a particular assessment. A baseline can help an organization understand starting points and design proportionate learning measures, but it is only one possible input.
Does AILAT certify EU AI Act Article 4 compliance?
No. AILAT measures aspects of AI literacy and supports programme planning. It does not classify AI systems, assess all legal duties, or certify compliance.
Is one general AI course enough?
Not necessarily. Appropriate measures depend on the people involved, the AI systems used, the operating context, and the associated opportunities and risks.
Does every employee need the same level of AI literacy?
No. Article 4 is context-sensitive, and the 2026 amendment (in force since 27 July 2026) clarifies that organizations are not required to guarantee a particular literacy level for each individual.
Can an AILAT report support internal documentation?
Yes. An aggregate baseline report can document the measured starting point and resulting learning priorities alongside the organization’s AI inventory, policies, training records, and other programme evidence.
Official sources and legal status
Regulation (EU) 2024/1689Regulation (EU) 2026/1744 (amendment)European Commission Q&A2026 amendment procedure
The Commission Q&A remains useful for practical guidance; where it describes the pre-amendment wording, the in-force text of Regulation (EU) 2026/1744 prevails.
Use evidence carefully
Establish the baseline, then design the programme around your real AI context.
AILAT helps your learning or governance team see the starting point. Your organization remains responsible for deciding what measures are appropriate.