{ "version": "2026-08-04", "licence": "CC BY 4.0", "note": "EU AI Act Article 4 (AI literacy) as amended by Regulation (EU) 2026/1744, with the original text for comparison, the role groups and evidence expectations behind the pk-sharma.com AI literacy checker. Article 4 text is reproduced from the Official Journal, which is freely reusable. Commentary, role groups and evidence lists are original.", "lastReviewed": "2026-08-04", "sources": [ { "label": "Regulation (EU) 2024/1689 (AI Act)", "url": "https://eur-lex.europa.eu/eli/reg/2024/1689/oj/eng" }, { "label": "Regulation (EU) 2026/1744 (Digital Omnibus on AI), amending Article 4", "url": "https://eur-lex.europa.eu/legal-content/EN/TXT/?uri=CELEX:32026R1744" } ], "article4": { "original": "Providers and deployers of AI systems shall take measures to ensure, to their best extent, a sufficient level of AI literacy of their staff and other persons dealing with the operation and use of AI systems on their behalf, taking into account their technical knowledge, experience, education and training and the context the AI systems are to be used in, and considering the persons or groups of persons on whom the AI systems are to be used.", "amended": "Providers and deployers of AI systems shall take measures to support the development of AI literacy of their staff and other persons dealing with the operation and use of AI systems on their behalf, taking into account their technical knowledge, experience, education and training and the context the AI systems are to be used in, and considering the persons or groups of persons on whom the AI systems are to be used. This obligation does not require providers or deployers to guarantee any specific level of AI literacy of any individual.", "amendedBy": "Regulation (EU) 2026/1744", "amendedInForce": "2026-07-27" }, "shifts": [ { "id": "means-not-result", "heading": "An obligation of means, not of result", "before": "Take measures to ensure, to your best extent, a sufficient level of AI literacy.", "after": "Take measures to support the development of AI literacy, with an express statement that you need not guarantee any specific level for any individual.", "soWhat": "You are now judged on what you provided, not on what each person absorbed. A programme that is offered, resourced and kept current is compliant even where an individual has not completed it. Under the old wording that gap was arguable; it is now expressly not a breach." }, { "id": "no-testing-mandate", "heading": "There was never a testing mandate, and now there is plainly none", "before": "Widely read as implying assessment, because 'a sufficient level' invites the question of how you would know.", "after": "No level to demonstrate, so no assessment is required to prove one.", "soWhat": "Competency testing, pass marks and certificates are a choice you may make for your own assurance. They are not what Article 4 asks for, and any vendor telling you otherwise is selling against a version of the text that no longer exists." }, { "id": "sme-support", "heading": "The Commission now has to help", "before": "No support duty.", "after": "The Commission and Member States must support and facilitate compliance, particularly for SMEs, and the Commission must publish practical examples of how to comply on the single information platform.", "soWhat": "If you are a smaller organisation, free official material is coming and is worth waiting for before commissioning anything expensive. The Omnibus also inserts definitions of SME and of 'small mid-cap enterprise' into the Act for the first time." }, { "id": "board-recommendations", "heading": "Recommendations are coming from the AI Board", "before": "Nothing specified.", "after": "The Board is to adopt recommendations on promoting AI literacy, taking European competence frameworks into account.", "soWhat": "A defensible programme should be able to absorb those recommendations later. Build something you can map onto a framework rather than a fixed curriculum you would have to rip out." } ], "keyDates": [ { "date": "2025-02-02", "label": "Article 4 became applicable", "detail": "AI literacy has been a live obligation since this date, alongside the prohibitions in Article 5. It did not begin in 2026, and it was never deferred." }, { "date": "2026-07-27", "label": "The Digital Omnibus entered into force", "detail": "Regulation (EU) 2026/1744 replaced Article 4 from this date. Any guidance written before it describes a version of the obligation that no longer applies.", "contested": true }, { "date": "2026-08-02", "label": "General application of the Act's governance and penalty regime", "detail": "The date from which the broader supervisory and penalty machinery applies. Article 4 itself was already in force; what changes around this point is the apparatus available to supervise it." }, { "date": "2026-12-02", "label": "New prohibitions apply", "detail": "The prohibitions inserted by the Omnibus into Article 5 apply from this date, rather than immediately on entry into force." }, { "date": "2027-12-02", "label": "Annex III high-risk obligations apply", "detail": "Deferred from 2 August 2026. Chapter III Sections 1 to 3 apply to systems classified high-risk under Article 6(2) and Annex III from this date." }, { "date": "2028-08-02", "label": "Annex I high-risk obligations apply", "detail": "Deferred from 2 August 2026. Chapter III Sections 1 to 3 apply to systems classified high-risk under Article 6(1) and Annex I from this date." } ], "penaltyPosition": { "headline": "Article 4 has no fine attached to it in the Act itself.", "body": [ "Article 99(4) sets out the infringements that attract administrative fines of up to EUR 15 million or 3% of worldwide turnover, and lists the provisions it applies to. Article 4 is not among them. There is no EU-level ceiling written against the AI literacy obligation.", "That does not make it unenforceable. Article 99(1) requires Member States to lay down rules on penalties for infringements of the Regulation and to make them effective, proportionate and dissuasive, so the exposure is whatever your Member State legislated, and it varies.", "One route does reach further. Where the AI Office supervises an operator directly, the Omnibus provides that infringement of any applicable provision of the Regulation, expressly including those not listed in Article 99(4), may attract fines under that Article. That is a real widening, and it applies to the AI Office's own supervisory perimeter rather than to every deployer." ], "misconception": "Marketing that quotes EUR 35 million or 7% against AI literacy is quoting the ceiling for the prohibited practices in Article 5, which is a different obligation entirely. Nothing in the Act attaches that figure, or the EUR 15 million figure, to failing to train your staff." }, "roleGroups": [ { "id": "everyone", "label": "Everyone with access to an AI tool", "who": "Anyone who can open an assistant, a copilot or an AI feature inside software you already licence. In most organisations this is now the whole workforce, whether or not anyone decided that.", "provide": [ "What the tool is, what it is for, and what it is not for.", "That output can be confidently wrong, and that confidence is not evidence.", "Which categories of information must never be pasted in, with examples rather than abstractions.", "How to raise a concern, and that raising one is welcome." ], "evidence": [ "The material itself, dated and versioned.", "A record that it was made available to the group, and when.", "The acceptable use policy it points at." ] }, { "id": "heavy-users", "label": "People whose work depends on AI output", "who": "Staff who draft, code, analyse, summarise or decide with AI assistance daily, where the output feeds something that matters.", "provide": [ "How to check output in their own domain, which is a different skill from general scepticism.", "Where the tool is systematically weak, taken from your own use rather than from a vendor's list.", "The boundary between assistance and delegation for their work.", "What to record about how a piece of work was produced." ], "evidence": [ "Role-specific material rather than the general briefing.", "Worked examples from your context.", "Whatever review step exists for AI-assisted work." ] }, { "id": "builders", "label": "People building or configuring AI systems", "who": "Engineers, data scientists, and anyone assembling agents, retrieval pipelines or automations, including through low-code tools.", "provide": [ "The failure modes of the architecture they are using, including prompt injection where the system reads untrusted content.", "What the organisation's obligations are for the systems they build, and which classification those systems fall into.", "Least privilege for anything the system can act with.", "The logging and traceability expected of them." ], "evidence": [ "Technical material and its currency.", "Design review records showing the considerations were applied.", "The system inventory entry naming who built and who owns it." ] }, { "id": "decision-makers", "label": "People who approve, procure or oversee", "who": "Executives, risk and procurement, and anyone signing off deployment or buying a system.", "provide": [ "Enough to interrogate a vendor claim rather than accept it.", "How classification works under the Act and what it obliges you to do.", "What questions to ask before approving a deployment.", "The organisation's own risk appetite, expressed in terms they can apply." ], "evidence": [ "Board or committee minutes recording the briefing.", "The procurement questions actually asked of a supplier.", "The approval record for a deployment." ] }, { "id": "affected", "label": "People the systems are used on", "who": "Not staff, and easy to miss. Article 4 requires you to consider the persons or groups on whom the systems are used, which for many organisations means customers, candidates or the public.", "provide": [ "Plain notice that a system is in use and what it does.", "How to reach a human, and how to contest an outcome.", "Nothing that requires them to be trained: this is about your design, not their education." ], "evidence": [ "The notice as published.", "The contest or appeal route as it actually operates." ] } ] }