Do we need to test our employees' AI knowledge under Article 4?
No. The European Commission's Q&A on AI literacy states that Article 4 does not entail an obligation to measure the knowledge of AI of employees. The duty is to take measures that support AI literacy, and the record of those measures is what matters. Regulation (EU) 2026/1744, in force from 27 July 2026, removed the outcome-style wording that had suggested otherwise.
Where the idea of testing came from
The original text asked for a sufficient level of AI literacy. A level sounds like something you measure, and a market of quizzes and certificates grew up around that reading. The rewritten article asks for measures that support literacy and says you are not required to guarantee any particular level for any individual. There is nothing left to measure against.
When a test is still useful
A short check at the end of a session can be a good teaching tool. It tells you which points landed and which need repeating, and it gives people a reason to pay attention. Used that way it is part of the measure. What it is not is evidence you are required to produce, and a low score is not a compliance failure.
What to record instead
Record what was covered, who attended, when, and why that depth suited those roles. That is the record the Commission describes, and it is what a regulator would ask to see.
Sources
- European Commission, AI literacy: questions and answers, the answer on whether there is an obligation to measure employees' knowledge.
- Regulation (EU) 2026/1744, the Digital Omnibus on AI.
This is general information about the EU AI Act, not legal advice. Enforcement is proportionate and case by case, and your sectoral regulator may issue guidance that applies to you.
