Training AI-literacy

Since the launch of ChatGPT, AI and algorithms have become an integral part of work. AI can be useful for quickly gathering information and for creating text and images. However, using AI also comes with risks, such as the risk of mistakes and even discrimination. organisations must therefore offer AI literacy training, that covers the risks of using AI and when to and when not to use AI. In this interactive training taking place over 2-2,5 hours, you will learn about the basics of AI, AI literacy, and the responsible use of AI.

Goal of the training

This training was developed with the professionals working in the Netherlands in mind. This training is useful for people who write or review texts, make use of software that might use AI, or people who use AI to improve digital processes. Prior knowledge of AI is not needed in order to take part: the training covers topics like the different kinds of AI systems, the rules concerning the usage of AI in accordance with the AI Act, and how to prevent mistakes by AI and algorithms.

Content of the training

The content of this training is based on the AI regulations in addition to the instructions of the Dutch data protection authority, the Autoriteit Persoonsgegevens (AP). The AP expects organisations to follow the following standards:

• The people involved with AI should have the knowledge required to recognise and mitigate the risks.
• The approach taken concerning AI should take into consideration the existing level of knowledge and experience of the people involved.

In the training, we will discuss a few examples in which AI was used and things went wrong. The following examples and risks will be discussed:

• Hallucinations: Information made-up by AI
• Nonsensical decisions made by AI
• Misleading pictures and deepfakes
• Bias and discrimination, for example, by using old data
• Unwanted actions, including hacks, by AI Agents
• Software development using AI
• Shadow-AI: software that users utilise themself without consulting the IT organisation.
• Lack of human oversight and control.
• Forbidden AI systems, like emotion recognition

Afterwards, all the participants will be aware of the risks and points of concern regarding the use of AI. Briefly, the topic of AI governance will be discussed: organisations will have to map the use of AI, in some cases using Fundamental Rights impact assessments and being in possession of a risk-management system.

Results

After completion of the training, every participant will receive a certificate. This makes them demonstrably trained in AI literacy. There will be more knowledge about the risks of AI, and people will have learned when to and when not to use AI, how to check the results, and recognise risks.

Costs of the training

The costs of the training depend on the level of tailoring to the organization and the number of participants. A standard training with no extra preparation, for 3-10 people, comes in at 800 Euro ex VAT.  A training tailored to the organization with advise regarding further steps comes in at 1600 Euro ex VAT.

Possible additional steps

This training is a good first step to make the organization fully compliant with the AI regulations and AVG legislation. We also provide other training courses, for example about the AVG and cybersecurity legislation. If needed, we are also capable of assisting organisations with setting up a risk-management system, either for AI in accordance with ISO 42001 or for information security in accordance with ISO 27001.

Source image: Blend Archive via Unsplash