Grow in Haus
AI Literacy Under the EU AI Act: How to Train Your Teams
Many organisations adopted generative AI tools faster than they trained anyone to use them. The EU AI Act addresses that gap directly. Article 4 of Regulation (EU) 2024/1689 requires providers and deployers of AI systems to take measures to ensure a sufficient level of AI literacy among their staff and others who use AI on their behalf. The obligation has applied since 2 February 2025.
The rules are being refined. Recent changes under the EU’s AI Omnibus recalibrate Article 4, but commentators are clear on the core point: AI literacy remains mandatory. What changes is the emphasis. Organisations are expected to take appropriate measures, rather than to guarantee that every individual reaches a defined level of competence, and the Commission and Member States are to support organisations with guidance and examples.
What “AI literacy” actually means
The Act describes AI literacy as the skills, knowledge and understanding that allow people to use AI systems in an informed way, to be aware of the opportunities and risks, and to understand the possible harm AI can cause. In day-to-day terms, that usually includes:
- Understanding what AI tools can and cannot do, including that they can produce confident but wrong answers.
- Knowing how to protect personal, confidential and client data when using AI tools.
- Recognising bias in outputs and in the data behind them.
- Keeping meaningful human oversight over decisions that affect people.
- Knowing when to escalate, and who owns what when something goes wrong.
Why one webinar will not be enough
Article 4 refers to staff’s technical knowledge, experience, education and training, and to the context in which AI is used. That points to a simple conclusion: the right level of AI literacy is different for different people. A recruiter screening CVs with an AI tool needs different knowledge from a marketer drafting copy, and both differ from the specialists who configure the systems.
Build a role-based learning path
- Take inventory. List the AI tools in use, officially and unofficially, and who uses them for what.
- Segment by role and risk. A common structure has three tiers: foundations for everyone, applied skills for regular users, and deeper training for people who select, configure or oversee AI systems and for managers making decisions based on AI output.
- Combine knowledge with practice. Short explainers help, but behaviour changes through use cases, exercises and feedback on real work.
- Set simple rules. Publish a short, readable AI-use policy: approved tools, data rules, review requirements, and where to ask questions.
- Keep a record. Document what training was offered, to whom and when. It helps you show you took appropriate measures.
- Refresh regularly. Tools and rules change quickly. Treat AI literacy as an ongoing programme, not a one-off event.
The human side most plans miss
AI literacy is not only technical. People also need the confidence to say “I don’t think this output is right” or “I’m not sure we should use AI for this.” That depends on psychological safety: a team climate where raising doubts is welcome. Without it, errors stay hidden until they become incidents.
Managers matter most here. They set the tone for whether people experiment responsibly, ask questions, or quietly work around the rules. Training for managers should cover how to discuss AI use openly, how to review AI-assisted work, and how to handle mistakes without blame.
Where People Haus can help
We design experience-based programmes that combine understanding with practice. Our digital transformation and technology skills training covers digital literacy, AI tools and change adaptation, and our leadership training prepares managers to guide teams through technology change. If you would like a plan tailored to your organisation and countries, talk to us.
This article is for general information only and is not legal advice. Rules differ by country and change quickly, so check current national law or ask your legal adviser before you act.
