AI literacy checklist for Article 4

How to use the AI literacy checklist

1. Enter your organisation and the person responsible (optional: they only appear on the record).
2. For each measure, choose Done, In progress or Not started.
3. Read your plan: the measures in progress first, then those not started.
4. Print the record, or choose "Save as PDF" in the print window. It carries today's date and each measure's status.

Your answers are kept in this browser, so you can come back and update them. "Start again" clears them.

How the checklist works (no AI model is called)

The checklist is a form that runs in your browser: no account, no server, no AI model. The record is printed by your browser's own print function. The measures fall into three groups:

• Know where AI is used: the tools in use, who uses them and for what, and the sensitive uses (personal data, confidential information, decisions about people).
• Take measures: a written AI use policy, the basics for everyone, hands-on practice with the tools actually used, the risks (confidential data, made-up answers, bias, prompt injection), deeper training for the roles that need it, human review of outputs, and a named contact.
• Keep it going, and keep a record: newcomers trained when they join, a refresh when tools change, and records of who was trained, when and on what.

What the result means

Article 4 of the AI Act (Regulation (EU) 2024/1689), as amended by Regulation (EU) 2026/1744, reads in the consolidated text in force on 27 July 2026:

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.

The original text asked them to "ensure, to their best extent, a sufficient level of AI literacy". Article 4 has applied since 2 February 2025 and is enforced since 2 August 2026 (the Commission's timeline). Our reading: what an organisation can show is what it did, the measures taken, for whom and when. That is what the record lists.

The record is a working document, not a certificate, and this page is not legal advice: no checklist can certify compliance, and the measures that fit depend on your tools, your staff and the risks of your uses. For a decision about your obligations, read the text and ask your legal adviser. Facts checked on 27 October 2026.

FAQ

Is the checklist free?

Yes, with no account. It runs entirely in your browser.

Are my answers sent anywhere?

No. They stay in this browser (in its local storage, when the browser allows it), and the record is printed by your browser. No AI model is called.

Does this checklist make us compliant with Article 4?

No checklist can. It helps you see which measures you have taken and which are left, and keep a dated record of them. What is enough for your organisation depends on your uses of AI.

Who does Article 4 apply to?

To providers and deployers of AI systems, for "their staff and other persons dealing with the operation and use of AI systems on their behalf". In practice, that includes most organisations whose staff use AI tools at work.

Is a training course enough?

A course is one measure. The checklist also covers knowing which tools are used, a policy, practice with the tools themselves, the risks, review of outputs and records, because those are what make training stick.

What kind of training counts as practice?

Practice means working on real tasks with the tools your staff use. Tokenacademy's sessions have each employee build with a real AI agent, with every milestone validated automatically and every token counted. Book a demo.