---
id: ai-act-article-4
lang: en
title: "Article 4 of the EU AI Act. What does AI literacy mean for your company?"
seoTitle: "AI Act Article 4: AI literacy explained"
description: "What does Article 4 of the EU AI Act ask of your company? Learn what AI literacy means, which measures you can take and how to record them for your team."
keywords:
  - AI Act article 4
  - AI literacy
  - AI Act AI literacy
  - EU AI Act for companies
  - AI literacy obligation
  - AI literacy training
  - AI Act SME
  - digital omnibus on AI
updated: "2026-11-13"
published: "2026-11-13"
# Linked from the "Build AI literacy" block on /en/pro/.
---

*Last checked on 13 November 2026*

Your customer service team uses Copilot to draft a reply. Your HR team has AI rewrite a job ad. Even everyday uses like these fall under **Article 4 of the EU AI Act**. If your company provides AI systems or uses them under its own authority, you must take measures to help develop the **AI literacy** of the people who work with them.

## What does Article 4 ask today?

Article 4 has applied since **2 February 2025**. The original text asked companies to ensure, to their best extent, a sufficient level of AI literacy. Since **27 July 2026**, the amended text from the **digital omnibus on AI** (Regulation (EU) 2026/1744) applies. It requires companies to support the development of AI literacy with concrete measures. Your company does not have to guarantee a set level of knowledge or skill per person.

When you choose measures, take into account the technical knowledge, experience and training of the people who work with AI, the way they use the system, and who may be affected by that use.

According to the European Commission, the rules for supervision and enforcement by national market surveillance authorities have applied since **2 August 2026**. The Commission and the member states support companies, especially SMEs. The Commission publishes examples of good practice and the European AI Board is tasked with drawing up recommendations.

## Who does Article 4 apply to?

- **Providers** develop an AI system or have one developed, and place it on the market or put it into service under their own name or trademark. Think of a software company that launches an AI assistant for its customers.
- **Deployers** use an AI system under their own authority. Think of an SME that uses ChatGPT for translations or Copilot in customer service. That use falls under Article 4 too. Purely personal, non-professional use falls outside this definition.

The measures cover your employees *and* other people who use or manage AI systems on your company's behalf. Think of a freelancer who writes product copy with AI for your company. Customers and job applicants can be affected by your AI use. Take those effects into account when you decide what your teams need to know and be able to do.

## What does AI literacy mean at work?

**AI literacy** is the knowledge and skill to work with AI consciously. Anyone who uses AI should be able to judge when a system helps, where it can make mistakes, which data may go into it, and when someone needs to check or step in. What that means in practice differs per role.

| Role | Example at work |
| --- | --- |
| **Customer service** | Copilot drafts a reply about a refund. The employee checks the amount and the terms in the customer file before sending. |
| **HR** | A recruiter has AI rewrite a job ad and checks that the requirements stay correct and inclusive. Does HR want to use AI to filter CVs or assess candidates? Then have someone check beforehand whether the separate rules for high-risk AI systems apply. |
| **Finance** | AI flags invoices that may have been booked twice. The employee compares the suggestion with the original invoices and the accounting system before making a correction. |
| **Management and team leads** | A team wants to put an AI chatbot on the website. The person in charge decides which questions the chatbot may answer, when an employee takes over, and who follows up on recurring mistakes. |
| **IT and application management** | An AI assistant gets access to internal documents. The administrator tests with different user roles which documents the assistant can retrieve, and records how colleagues report unwanted access. |

## Which measures can you take and record?

Start with the tools your teams use today. A short onboarding with clear agreements on tools, data and checks can give you a base. Add exercises per role. For example, have customer service review an AI reply that contains a made-up refund condition. Give people a contact person for doubts or mistakes, and update the instructions when tools or tasks change.

**Keep a simple internal overview.** For each use, note who works with it, who may be affected, which risks you saw and why you chose certain measures. Also keep the date, audience and content of trainings or instructions, who took part or which teams you reached, and who handles the follow-up. Article 4 does not prescribe a fixed template for this.

## What role do training and certificates play?

Training helps employees practise with the AI tools and situations they meet in their work. Article 4 prescribes **no fixed training format, exam or certificate**. A certificate of completion can show who followed which training. Use it as part of your records, together with the agreements, exercises and follow-up that fit your own AI use. On its own, a certificate does not guarantee that your company complies with Article 4.

## Checklist to get started

- [ ] List the AI tools your teams use and what they use them for.
- [ ] For each use, note who works with it and who may be affected.
- [ ] Decide per role what people already know, what they need to learn and which checks they carry out.
- [ ] Record the agreements, practise with familiar situations and name a contact person.
- [ ] Keep a record of what you did, and review the measures when tools, tasks or recurring mistakes change.

## Build AI literacy with Rollo Academy Pro

Give your team practical AI training and keep track of their progress. That gives AI literacy a concrete place in your training policy. [Discover Rollo Academy Pro](/en/pro/).

## Official sources

- [EUR-Lex · EU AI Act, consolidated English text of 27 July 2026. See Articles 2, 3, 4, 26 and 113 and Annex III.](https://eur-lex.europa.eu/legal-content/EN/TXT/?uri=CELEX%3A02024R1689-20260727)
- [EUR-Lex · Regulation (EU) 2026/1744 in the Official Journal. Article 1, point 5 replaces Article 4 of the AI Act.](https://eur-lex.europa.eu/legal-content/EN/TXT/?uri=CELEX%3A32026R1744)
- [European Commission · Article 4 in the AI Act Service Desk.](https://ai-act-service-desk.ec.europa.eu/en/ai-act/article-4)
- [European Commission · AI literacy questions and answers on training, documentation and supervision.](https://digital-strategy.ec.europa.eu/en/faqs/ai-literacy-questions-answers)
- [European Commission · The digital omnibus on AI enters into force.](https://digital-strategy.ec.europa.eu/en/news/ai-omnibus-enters-force)
