EU AI Act: Are AI training now mandatory?

The EU AI Act is in force and with it a decisive innovation for companies. The duty to "AI Literacy".

Since 2 February 2025, all companies using artificial intelligence must ensure that their employees have sufficient AI competencies. But what does that mean for your company?

Are formal AI trainings now a must? This article gives you a complete overview of the new legal situation and shows how to pragmatically implement the needs.

What exactly does the EU AI Act require in Article 4?

The AI Act of the European Union, officially adopted as Regulation (EU) 2024/1689, aims to create a uniform legal framework for the development and application of artificial intelligence.

A central component of this rule is Article 4, which focuses on "AI literacy".

"Suppliers and users of AI systems take measures to ensure a sufficient level of AI competence for their staff and other persons. "

This provision means that companies must become active. It is no longer enough to implement AI systems.

They must also ensure that all employees who come directly or indirectly into contact with these systems understand their functioning, options and risks.

This includes technical, legal and ethical aspects.

Why is AI competence so crucial?

EU AI Act Training Obligation 2026: Learn whether AI training is mandatory for your company and how you meet the needs of Article 4.

For EU AI Act: Are AI training now mandatory? , see Cost Calculator: AI Development und Discover solutions on our website for rollout paths and planning.

The intention of the legislator is clear. A responsible and effective use of AI is only possible if users have the needed knowledge.

Without this understanding, major risks can arise, from data breaches, discriminatory decisions, and reputational damage.

Well-trained employees are the first line of defence against such risks and a decisive factor for the successful and value-generating integration of AI into corporate processes.

Are formal training inevitable?

Short: The good news first: The AI Act does not explicitly require external, certified training.

The good news first: The AI Act does not explicitly require external, certified training. Companies have a certain scope to ensure their employees' AI competence.

IHK Schleswig-Holstein emphasizes that ownership lies first in the company's own ownership.

However, the rollout of structured training measures is the most obvious and secure way to comply with the care requirement. The documentation of these measures is crucial.

The aim is to be able to show in the event of an audit or a case of damage that the company has complied with its obligation.

Action recommendations for your company

The industry association Bitkom recommends a structured approach to meet the needs of the AI Act:

Step Measure Description

1 Death review Determine which AI systems are in use and which employees work with them. Bewer

---

Method note: External statistics refer to published industry and official data (Bitkom, Destatis) where not otherwise attributed. Company-specific figures: Groenewold IT, 2026.

References and further reading

The following separate references complement the topics in this article:

• Bitkom – German digital industry association.
• German Federal Office for Information Security (BSI).
• European Commission – Digital strategy.
• MDN Web Docs (Mozilla)
• W3C – World Wide Web Consortium.

"Legacy migration often fails not because of the stack, but because tacit domain knowledge was never captured—budget explicitly for knowledge transfer."

— Björn Groenewold, Managing Director, Groenewold IT Solutions

Frequently Asked Questions (FAQ)

What is this article about: “EU AI Act: Are AI training now mandatory?”?

This article sums up practical aspects of EU AI Act. Are AI training now mandatory? for leaders and delivery teams. In short. EU AI Act Training Obligation 2026.

Learn whether AI training is mandatory for your company and how you meet the needs of Article 4.

Who benefits most from the content described here?

It is especially relevant for firms in AI training that need reliable systems, clear interfaces, and predictable delivery — from mid-market teams to expert departments.

How does this topic fit into an IT or digital strategy?

You can map the topic to service building blocks such as custom software and delivery support. Architecture reviews and stepwise rollout reduce risk and rework.

For multi-system landscapes, IT consulting and architecture helps align vendors and internal teams.

What are sensible next steps if we need support?

For architecture, rollout, or a second expert opinion, book a free initial consultation — including timeline and interface alignment.