6 · Train

Train your teams on AI, and be able to show it.

Article 4 of the Act requires that the people who use or oversee an AI system have a sufficient level of AI literacy. We build paths by audience, assessed before and after, and keep the register of attestations you will be able to produce.

Executives: Half-day session
Business teams: Modules of 1 to 2 hours
Format: Paths by audience, one-off sessions or cycles
Evidence: Register of attestations

Build a programme

A duty that covers every system

The AI literacy requirement does not turn on the risk tier. It applies to a writing assistant as much as to a scoring model, and it binds providers and deployers alike. What counts as sufficient depends on each person’s role and on the context of use.

Training has to be evidenced

An attendance sheet does not demonstrate a level of literacy. We assess skills before and after each path, and every participant receives an attestation recorded in a central register you can put in front of an auditor.

Scope

Four audiences, content drawn from your own cases.

An executive, a business user, a developer and a lawyer do not need the same training. Each path starts from the tools that audience actually uses.

Executives. A half-day AI strategy session, followed by a briefing on the legal obligations that fall on the organisation.

Business teams. E-learning modules of 1 to 2 hours by profile, and hands-on prompting practice on the tools in service.

Data and engineering teams. A module on AI risk and security, responsible-build playbooks and a review checklist for a model or an agent.

Legal and compliance. A module that follows the Act article by article, with what each obligation means in practice.

Sector case studies. A library of use cases drawn from your sector, worked through in session.

Applied coaching. Follow-up after the training, so that what was learnt carries into daily work.

AI usage policy. A policy template to adapt, setting out what is allowed, what is supervised and what is excluded.

Network of champions. An internal network of AI champions and a twelve-month roadmap to sustain the level reached.

Who trains

Practitioners. Sessions are led by people who build and audit AI systems themselves, working from your real cases.

How it runs

Six steps, from audience mapping to the register.

Audience mappingWho uses an AI system, who oversees it, who decides. It draws on the system inventory where one exists.
Baseline assessmentThe starting level of each audience, measured before any session, to size the paths.
Path designContent built on your tools, your use cases and your AI usage policy.
Sessions and modulesThe executive session, the e-learning modules, the hands-on workshops and the specialist modules, as one-off sessions or in cycles.
Final assessment and attestationsThe level reached is measured, each participant receives an attestation and the central register is updated.
Sustaining the levelThe twelve-month roadmap, the network of champions and periodic reporting on progress.

Deliverables

What you receive.

Executive session Paths by audience Sector case studies AI risk and security module Responsible-build playbooks Article-by-article regulatory module Before-and-after assessment AI literacy attestations Central register of attestations AI usage policy template Twelve-month roadmap Network of AI champions

Who the training is for

HR and learning functions, compliance functions and AI governance leads who have to show that teams hold the level required to use and oversee the systems in service.

Before and after the training

Training is what makes the framework live.

Before · Structure

AI governance framework

The policies, roles and decision paths the training brings to the teams.

After · Deploy

AI delivery

AI product owners and AI project directors to take your projects through to production.

Next step

Build a programme.

Tell us which audiences are in scope and which AI tools they use. We will come back with a programme, a format and a fixed fee.

Write to us