How to Plan AI Literacy Required by the AI Act

01

What the AI Act requires regarding AI literacy

Article 4 of Regulation (EU) 2024/1689 requires providers and deployers of AI systems to take measures to ensure, to their best extent, a sufficient level of AI literacy among their staff and any other persons dealing with the operation and use of AI systems on their behalf. Planning must consider the technical knowledge, experience, education and training of those using the systems, the context of use, and the persons or groups on whom the systems are used.

02

Key definitions to start from

AI literacy is defined as the skills, knowledge and understanding that allow providers, deployers and affected persons to make an informed deployment of AI systems, while becoming aware of the opportunities, risks and possible harms it can cause. The provider is the one who develops or has developed an AI system and places it on the market; the deployer is the one who uses it under their own professional authority.

03

How to build the training plan

Steps

1. Map the AI systems in use and identify who operates or uses them on the company's behalf.
2. Assess the technical knowledge, experience, education and training of the people involved.
3. Consider the context of use and the persons or groups on whom the systems act.
4. Link the training paths to the risk management system for high-risk systems.
5. Periodically review and update the training, following the iterative risk management process.

04

Hypothetical example: AI-assisted personnel selection

Hypothetical example: a company using an AI system to select candidates acts as a deployer. The team that configures the system and the HR staff who interpret its outputs have different training needs: the former must understand technical settings, the latter the context of use and the impact on the people being evaluated, as required by the regulation's literacy criteria.

05

Minimum checks before launching training

• List of AI systems and the roles, provider or deployer, associated with each.
• Record of the technical knowledge and experience of the people involved.
• Note on the operational context and the persons targeted by the system.
• Link between training and the stages of the risk management system.
• Evidence of periodic review of the training plan.

06

Setting up a documentation trail with a software partner

In the absence of verified product features, a company may choose to configure, with its software partner, a register of AI systems indicating role, people involved and the level of knowledge required, so as to make the training planning required by Article 4 traceable. It may also plan a periodic review of this register, linked to the identification and update stages provided for in the risk management system for high-risk systems.

07

Clarifications on the scope of the obligation

The definition of AI literacy covers both technical skills and understanding of risks and opportunities, not just operational use of the system.

The Commission may approve codes of good practice or, if not adequate, adopt common rules for the implementation of obligations such as those on literacy.

08

The practical next step

The practical next step is to build an inventory of AI systems with roles, required skills and a link to the risk management system, to be reviewed periodically.