The EU AI Act and AI literacy: what it requires of you
Since 2 February 2025, companies in the EU have a legal obligation to ensure their staff understand the AI tools they use. Here is what that means in practice.
RegulationPublished 8 min read
Regulation (EU) 2024/1689, known as the AI Act, is the world's first comprehensive legal framework for artificial intelligence. It entered into force on 1 August 2024 and applies in stages over the following years.
For most companies the most relevant part is not the one on high-risk systems. It is a short article that already applies and that affects every employer whose staff use AI tools.
A deployer is any organisation using an AI system in its activity. If your team uses an AI tool for work, you are a deployer.
What this means in practice
The obligation has applied since 2 February 2025. It does not require a certificate, an exam or a specific course. It requires the employer to have taken reasonable measures so that the people working with AI understand what they are doing.
The Regulation defines AI literacy as the skills, knowledge and understanding that allow informed deployment of AI systems, as well as awareness of the opportunities, risks and possible harm. The level required is judged against people's technical knowledge, experience and education, and the context of use.
What AI can do
People need to understand which tasks the tool suits and where it genuinely helps in their work.
What it cannot do
Limitations are part of literacy: that a model can be confidently wrong, that it does not know what it does not know, and that output always needs checking.
What the risks are
Data privacy, sensitive information in prompts, bias in output, and responsibility for anything that reaches a customer or a decision.
Dates worth knowing
1 Aug 2024
Entry into force
The clock starts on the staged application of each part.
2 Feb 2025
AI literacy and prohibited practices
Article 4 applies. At the same time, prohibitions take effect on practices such as social scoring and emotion recognition in the workplace.
2 Aug 2025
General-purpose AI models
Obligations for providers of general-purpose models apply, along with governance and penalty rules.
2 Aug 2026
General application
The bulk of the Regulation applies, including requirements for high-risk systems under Annex III.
What to do
- Inventory which AI tools are already in use and by whom. There are usually more than you expect.
- Set out in writing which data may and may not go into those tools.
- Provide role-appropriate training: someone in finance needs different understanding from someone in marketing.
- Record what you have done. The obligation is to have taken measures, so documentation matters.
- Assign ownership and set a review cadence.
Why this is part of our programme
We are not a law firm and this article is not legal advice. But the literacy obligation lines up almost word for word with what the programme does anyway: people understand what the tool is for, where it fails, what must not go into it and who is responsible for the result.
That is why the programme has a module on checking and boundaries, and why the platform keeps a record of who went through what and what they applied. It is useful for the adoption itself, and the same record does the job when you need to show that measures were taken.
Topics
- EU AI Act
- AI literacy
- Article 4 AI Act
- AI regulation
- employer obligations AI
- Regulation 2024/1689
Want the same at your company?
Twenty minutes is enough to work out where it makes sense to start and what can actually be measured.