In this article (8 sections)
The AI Act obligation that applied earliest and has the widest scope is not the list of prohibited practices but AI literacy in Article 4. It has applied since 2 February 2025 to providers and deployers of all AI systems, regardless of risk level. The Digital Omnibus rewrote this article. Many people read that as “it has been relaxed”, but it is more accurate to say that the nature of the obligation has changed while the obligation itself remains.
Key takeaways
- Article 4 applies to providers and deployers of all AI systems, including companies that only use generative AI tools in the office.
- Before the Omnibus: “take measures to ensure, to their best extent” a sufficient level of AI literacy among staff. After the Omnibus: “take measures to support the development” of AI literacy, with an explicit statement that no specific level has to be guaranteed for any individual.
- New paragraph 2: the Commission and Member States must support companies (especially SMEs) in meeting the obligation, and the Commission must publish practical examples on the single information platform.
- Deployers of high-risk AI are also subject to Article 26(2): persons assigned to human oversight must have the necessary competence, training and authority, which is a more specific requirement.
What is AI literacy?
The definition in Article 3(56): the skills, knowledge and understanding that allow providers, deployers and affected persons, taking into account their respective rights and obligations under this Regulation, to make an informed deployment of AI systems, as well as to gain awareness about the opportunities and risks of AI and possible harm it can cause.
The factors Article 4 asks companies to take into account include:
- the technical knowledge, experience, education and training of staff and other persons dealing with the operation and use of AI systems on the company’s behalf;
- the context in which the AI systems are used;
- the persons or groups of persons on whom the AI systems are to be used.
So AI literacy is not one standard course. It has to match “who uses which AI, on whom”.
What did the Omnibus change?
| Before the Omnibus | After the Omnibus (from 27 July 2026) | |
|---|---|---|
| Core wording | take measures to ensure, to their best extent, a sufficient level of AI literacy | take measures to support the development of AI literacy |
| Required outcome | Ensure a sufficient level | Explicitly states that no specific level has to be guaranteed for any individual |
| Role of government | Not specified | Paragraph 2: the Commission and Member States support companies, especially SMEs; the Commission publishes practical examples |
| Common objectives | Not specified | Paragraph 3: the Board makes recommendations with reference to European competence frameworks, to set common objectives |
The key change is a shift from an “obligation of result” to an “obligation of conduct”. Companies do not have to prove that every employee has reached a certain level, but they must be able to explain which measures they have taken.
Article 4 applies in all of these cases
- Staff use generative AI to write emails, translate or make presentations
- Customer service, sales or HR use AI tools in their work
- You develop or sell products with AI functions
Article 4 applies regardless of risk level and is not limited to companies in the EU: as long as you are a provider or deployer and the AI Act applies, you have this obligation.
After the relaxation, what do companies still have to do?
The provision does not prescribe a specific format, number of hours or certificate, but “taking measures” means there must be actions you can explain. For a company, Article 4 involves at least three questions:
- Who: which employees and contractors operate or use AI systems?
- What: which opportunities, risks and possible harms do they need to know about? Does this match the systems they use and the context they use them in?
- Records: if a customer, auditor or authority asks, can you explain which measures you have taken?
The Commission will publish practical examples on the AI Act single information platform, which you can follow on the AI Act Service Desk.
How does AI literacy differ from human oversight of high-risk AI?
| Article 4 AI literacy | Article 26(2) human oversight | |
|---|---|---|
| Applies to | Providers and deployers of all AI systems | Deployers of high-risk AI |
| Requirement | Take measures to support the development of AI literacy | Assign oversight to natural persons who have the necessary competence, training and authority, and the necessary support |
| When it applies | 2 February 2025 (Omnibus version from 27 July 2026) | Annex III: 2 December 2027; Annex I: 2 August 2028 |
What the Omnibus relaxed is Article 4. For companies using high-risk AI, the competence and training of the people assigned to oversight remain a specific requirement, and the more relaxed version of Article 4 cannot be used in its place.
Can you be fined for breaching the AI literacy obligation?
The list of fines in Article 99(4) (up to EUR 15 million or 3%) does not include Article 4. However, Article 99(1) requires each Member State to lay down penalties and other enforcement measures for infringements of the Regulation, which may include warnings and non-monetary measures. National approaches are still developing.
The more practical pressure comes from the market: supplier questionnaires and audits from European customers have already started asking “do you have an AI use policy and staff training?”
Teach the AI Act to your team
Different departments use different AI and need to know about different risks. The content and depth that sales, HR, R&D and customer service need depends on the tools they actually use and the people they use them on.
Bring a list of the AI tools each department uses and book 30 minutes to discuss in-house training →
FAQ
Has the AI literacy obligation been removed?
No. The Omnibus changed Article 4 to “take measures to support the development of AI literacy” and states that no specific level has to be guaranteed, but providers and deployers still have an obligation to take measures.
If employees are just using ChatGPT on their own, is the company responsible?
Article 4 covers persons who operate or use AI systems “on behalf of” the provider or deployer. When employees use AI tools in their work, the company is the deployer.
Is there an official AI literacy course or certificate?
The AI Act does not prescribe any specific course or certificate. Article 4(2) requires the Commission to publish practical examples on the single information platform, and the Board will also make recommendations under paragraph 3.
Back to the start of the series
Part 1: The new AI Act timeline after the Digital Omnibus
Previous: Part 9: “We are not high-risk”. What evidence does the Article 6(3) exception require?
Need help?
ZC Zeelandia Consulting has an office in Amsterdam and provides in-house training and AI Act applicability checks, helping teams understand the AI rules that relate to their own work.
- Book a 30-minute online consultation: calendly.com/zc-meeting/30min
- EU Regulatory Radar: AI Act page
- Follow ZC on LINE: lin.ee/QKV5RTE
Legal sources and references
- Regulation (EU) 2024/1689 (Artificial Intelligence Act), Article 3(56), Articles 4, 26, 99 and 113: EUR-Lex
- Regulation (EU) 2026/1744 (Digital Omnibus on AI): EUR-Lex
- AI Act consolidated text (27 July 2026): EUR-Lex
- European Commission, AI Act Service Desk: ai-act-service-desk.ec.europa.eu
This article provides general information, updated to November 2026, and does not constitute legal advice. Whether a specific product is affected depends on the actual facts and the latest Official Journal texts.
Next step: have you considered these three options?
Sources · Regulation (EU) 2024/1689 (Artificial Intelligence Act) · Regulation (EU) 2026/1744 (Digital Omnibus on AI) · AI Act consolidated text (27 July 2026) · European Commission, AI Act Service Desk. This article is general information and not legal advice for any specific case. Obligations are defined by the Official Journal of the EU.
AI ActAI literacyArticle 4employee trainingDigital Omnibushuman oversight2024/1689
