One month in: meeting the AI Literacy requirements under the AI Act

11/03/2025 - Since February 2, AI literacy is no longer just a ‘nice-to-have’ but a mandatory requirement under the AI Act for organisations that develop, deploy, or use AI systems. This means that everyone in the AI value chain – from developers to end-users – must be able to understand how AI works, the associated risks, and how to apply AI systems responsibly.

In our blog post from November 20, we discussed the importance of AI literacy and the factors that influence it. With the Dutch Data Protection Authority (AP) now publishing a guidance document, this post takes a closer look at how organisations can meet the obligation to promote AI literacy among their employees. It will cover what AI literacy entails, how the AP recommends addressing this systematically, and additional measures organisations can take to implement these actions.

What is AI Literacy?

AI literacy, as outlined in Article 4 of the AI Act, requires providers and users of AI systems to ensure that their personnel and other relevant parties possess sufficient AI skills, taking into account their knowledge, experience, training, and the context in which the systems are used. This goes beyond technical expertise; it also includes an awareness of the ethical and social implications, such as bias, and the impact of AI-driven decision-making. Employees must be capable of understanding AI systems, critically interpreting their outputs, and applying AI responsibly within their roles.

A solid foundation in AI literacy enables organisations to implement AI systems effectively and in compliance with regulations. It helps prevent unintended discrimination and ensures human oversight of AI decision-making. The ‘human-in-the-loop’ approach, where individuals can interpret AI outputs and remain involved in critical decisions, plays a crucial role in this.

The AI Act specifies that the required level of AI literacy varies depending on the context, such as the sector, the phase of the AI application (development, testing, or deployment), and the roles of those involved. For example, AI used in healthcare demands different knowledge and skills than AI in the financial sector.

The Dutch Data Protection Authority's Four-Step Plan

As the authority responsible for supervising AI and algorithms, the Dutch Data Protection Authority (AP) has developed a guidance document to assist organisations in building AI literacy among their employees. The document proposes a multi-year action plan consisting of four steps. A summary of these steps is provided below.

1. Identification

   Organisations should identify which AI systems are in use and determine which employees interact with these systems. It's also essential to assess the current level of knowledge among employees—this could be done through surveys—and analyse the risks associated with the use of these AI systems.

2. Goal setting

   Based on the knowledge inventory and risk assessment, organisations need to define the knowledge and skills required for each role to use the AI system responsibly. For instance, employees working directly with an AI system should have (technical) knowledge about the system’s operation and outputs, while those who don’t interact with the system should only be aware of its existence and purpose.

3. Implementation

   To promote AI literacy, organisations must organise targeted training tailored to the specific needs of employees. Additionally, AI policies should be implemented, and a point of contact, such as an AI officer or working group, should be appointed to oversee the responsible use of AI.

4. Evaluation

   AI literacy requires ongoing attention and adaptation. Therefore, organisations should regularly monitor and evaluate progress. Policies and training materials should be updated promptly in response to new developments and regulations.

The AP emphasises that AI literacy is essential for both individuals and organisations to apply AI responsibly. This requires customisation, strategic planning, and continuous adjustments. According to the AP, organisations must actively invest in building knowledge and raising awareness to comply with legal requirements and manage the social impact of AI.

AI Literacy in Practice

Point 3 mentions the importance of providing targeted training based on the desired level of knowledge for employees. This could include a combination of:

• E-learning modules for broad, foundational AI knowledge across the organisation.
• In-depth workshops for employees working directly with AI.
• Practical, case-oriented sessions to address specific use cases and ethical issues within the organisation.

What’s Next?

At Considerati, we understand how complex it can be to comply with the new AI regulations. That’s why we offer customised AI literacy and AI governance training. We can also deliver an AI-officer, help develop AI policies and conduct risk analyses to ensure AI is used responsibly.