AIM SHIELD Emergency
AI training · Free

Your team uses AI. Teach them to use it carefully, in twenty minutes.

A six-block session for Article 4 of the European AI Regulation, an internal record that documents it and a one-page use policy. Everything is prepared in your browser: nothing reaches us.

What the law requires
If AI tools are used in your company, even other people's (a chatbot, a translator, an email assistant), you have to take measures to help your team understand what they do, their risks and how to avoid them. The obligation has existed since 2 February 2025.
What it does not ask for
Since the July 2026 reform, no specific level is required and, according to the European Commission, no certificate is needed either. What it asks for is measures suited to your work, to what each person knows and to the risks of the tools.
Why keep a register
The European Commission states that you can keep an internal record of training. It is a simple way of showing what you have done.

The session, in six blocks

About three minutes per block. Tell it with examples from your own work and leave time for questions.

  1. What the AI we use is and is not. It predicts text or images from examples: it does not know whether something is true, does not know your company except for what you give it, and can answer with great confidence something that is false.
  2. Check before using. Everything that comes out of an AI is checked before sending it, publishing it or deciding anything with it: figures, names, laws, quotations and links. If it can't be checked, it isn't used.
  3. What you should never paste. Personal data of customers or employees (ID numbers, accounts, health, addresses), passwords, confidential information and anything not yet published. If necessary, they are replaced with labels such as [CLIENTE] or [IMPORTE].
  4. Which tools are allowed. Only those the company has approved, with the company account. No uploading work documents to personal tools.
  5. Scams made with AI. Cloned voices asking for an urgent transfer, manipulated videos and emails without mistakes that imitate a supplier. When there is a request for money or data, it is confirmed by another route using the usual number.
  6. Who decides and who is told. AI helps, but a person decides and answers for whatever is sent. A mistake, a data leak or a deception is reported straight away to the person responsible.

To practise spotting scams, Train your eye. For a doubtful message, Check.

Prepare the record and the policy

Blocks delivered

None of this leaves your browser. If you close the page, it is lost: download or print the documents.

AI training record

AI use policy

Frequently asked questions

Does it affect me if we only use a chatbot?

Yes. Article 4 applies to anyone who uses AI systems at work (the regulation calls them a “deployer”), even if the tool belongs to another company.

Do I need an official certificate or course?

No. The European Commission states that no certificate is needed. What counts is measures adapted to your team; this record documents them.

How often does it need repeating?

The law does not set a deadline. Our proposal: once a year, when someone joins, and when you start using a new tool.

What if we use high-risk AI (for example, to select staff)?

Then there are additional obligations, such as training whoever oversees the system. This session is the foundation; it does not replace them: check with whoever advises you.

Is it stored anywhere?

No. Everything is written in your browser and nothing reaches us. Download or print the register and the policy before closing the page.

Official sources

Consulted on 29 October 2026. This is not legal advice.