Knowledge base
What is AI literacy? Meaning and examples per role
By Bob Biesemans · published on
AI literacy means knowing what AI can and cannot do, and using it safely and sensibly at work. The meaning under the AI Act, with examples per role.
Short answer
AI literacy is knowing what AI can and cannot do and using it in an informed way, aware of the opportunities, risks and possible harm. It is not about programming, but about things like spotting a wrong AI answer, keeping sensitive data out of public tools and staying responsible for the result.
The meaning in one sentence
AI literacy is the skills, knowledge and understanding that let people use AI systems in an informed way and be aware of the opportunities, the risks and the possible harm. That is how Article 3 of the AI Act defines it.
In plain words: you know what an AI tool does well and badly, you notice when an answer is wrong, you know which data must not go into it, and you stay responsible for what you do with it.
What it is not
AI literacy is not programming and not a technical course. An accountant does not need to know how a language model is trained, but does need to see that an AI summary of an invoice copied an amount wrongly.
Nor is it the same as using AI often. Opening a chatbot every day does not mean someone recognises a wrong answer or writes a good prompt.
Examples per role
Administration: reading an AI meeting summary before it is shared, and not pasting colleagues' names or health data into a public tool.
Sales: checking an AI-written quote email for wrong prices and promises, and using customer data only in the tools the company allows.
HR: knowing that an AI tool ranking CVs can repeat bias, and that such a decision is always taken by a person.
Logistics and production: checking an AI schedule for exceptions the tool does not know, such as a customer who only accepts morning deliveries.
What the AI Act asks
Article 4 asks organisations that use AI systems to take measures to support the AI literacy of their staff, tailored to their knowledge, experience and the context of use. Since the Digital Omnibus (Regulation (EU) 2026/1744) this is an obligation of means.
There is no mandatory certificate and no fixed training length. What counts is being able to show which measures you took, for whom and when. This is general information, not legal advice.
How to measure AI literacy
Asking whether someone knows the rules says little. AI literacy shows when a person does a task from their own work: judging an AI answer, spotting sensitive data, improving a prompt.
That is why a practical test works better than a quiz. Measuring the same people again after training also shows whether the measure worked.
Source: AI Act, Verordening (EU) 2024/1689, artikel 3 en 4 (EUR-Lex), Europese Commissie, AI literacy Q&A
Frequently asked questions
Is AI literacy mandatory?
Article 4 of the AI Act asks organisations that use AI systems to take measures to support their staff's AI literacy. Since the Digital Omnibus this is an obligation of means, with no mandatory certificate or fixed training length.
What is the difference between AI literacy and AI skills?
AI literacy is the baseline everyone who works with AI needs: understanding, checking critically, handling data safely. AI skills go further, such as writing a good prompt or setting up an AI workflow. Both are best measured with tasks from real work.
How do you measure employees' AI literacy?
With a practical test in which employees do tasks from their own work, such as judging an AI answer or spotting sensitive data. A second measurement after training shows whether the measure worked.