Pegamento’s AI literacy test isn’t meant to be a simple introductory quiz, but rather a serious knowledge check for professionals who want to understand and apply AI responsibly. And yes, some of the questions are quite challenging. That’s by design.
AI is increasingly being used in communications, healthcare, customer service, education, government, and business processes. As a result, it’s no longer enough to simply know what ChatGPT is or how to write a good prompt. It’s just as important to understand what AI systems can and cannot do, what risks are involved, and when human oversight remains necessary.
This test will help you discover the breadth of your AI literacy: from generative AI, language models, and AI agents to privacy, bias, reliability, and responsible use.
The test consists of 25 multiple-choice questions, divided into two sections.
Part 1: AI Knowledge and Responsible Use — Questions 1 through 20
This covers topics such as:
Generative AI and Large Language Models (LLMs): What They Are, How They Work, and How They Differ from Other Forms of AI.
Embeddings, vector databases, and RAG: techniques that enable AI to find and use relevant information from documents and data sources.
Prompt and Context Engineering: How to Give AI Good Instructions and Provide the Right Context and Information.
Agentic AI: AI systems that can plan tasks, use resources, and carry out actions within defined parameters.
AI Reliability: Topics include hallucinations, source verification, and the importance of human oversight.
Privacy and security: handling personal data and confidential information with care, prompt injection, and considerations regarding open-source AI models.
Bias and Discrimination: Identifying Risks When AI Is Used in Areas Such as Recruitment, Evaluation, and Decision-Making.
Part 2: Governance, Legislation, and Responsible Use of AI — Questions 21 through 25
In this context, we will examine, among other things:
human oversight and accountability in AI applications;
AI literacy and the requirements of the European AI Act;
data governance, including the origin, quality, relevance, and potential bias of data;
responsible use of AI in situations where the consequences for people may be more significant;
Copyright and Other Legal Considerations Regarding Generative AI.
Before you begin the test, we ask you to fill in your first name, last name and e-mail address.
You may also fill in your company name and position, if you wish. These details are needed to:
your score in person,
Offer you a retake if you score below 80%,
AND, with a perfect score, automatically generate and send your certificate.
We use your data exclusively for this purpose. They are processed securely and not shared with third parties.
Could you use some help?
Below, we’ll explain a few terms you’ll come across during the test. Of course, we won’t give away all the answers.
What is a Large Language Model?
A Large Language Model (LLM) is an AI model that has been trained on large amounts of text and can recognize and use patterns in language to generate new text. LLMs are used, for example, to answer questions, write texts, and summarize documents.
What is a vector embedding?
An embedding is a numerical representation of, for example, a word, sentence, or document. This allows computers to determine which pieces of information are similar in content. Embeddings are used, among other things, for semantic search and retrieving relevant information.
What is RAG?
RAG stands for Retrieval-Augmented Generation. In this process, a language model is provided with relevant information from external sources, such as documents or a knowledge base, before it generates a response. This allows AI to base its responses on information that does not come solely from the language model itself.
What are hallucinations in AI?
An AI model can sometimes provide an answer that sounds convincing but is actually incorrect or completely made up. We call this a hallucination. That is why it remains important to critically evaluate AI responses.
What are zero-shot and few-shot prompting?
In zero-shot prompting, you give an AI model a task without providing any examples. In few-shot prompting, on the other hand, you provide one or more examples to make it clearer what kind of response you expect.
What is Agentic AI?
Agentic AI refers to AI systems that can plan and execute multiple steps within specific goals and frameworks. Different AI agents or processes can also collaborate in this process. The organization and coordination of these agents is called agent orchestration.
And technology isn’t the only thing that matters
AI literacy also means being able to identify risks. Consider privacy, personal data, bias, discrimination, copyright, the reliability of AI output, and the question of when human oversight remains necessary.
Why This Test Is Valuable
This test isn’t just a simple quiz. It helps you discover just how broad your knowledge of AI has become: from LLMs, prompting, and RAG to data, risks, and responsible use.
Whether you work in healthcare, education, government, or the private sector, anyone who understands how AI works—and where its limitations lie—can make better decisions about how to use it.
If you achieve a perfect score of 100%, you’ll receive a certificate from Pegamento in recognition of your knowledge. If your score is lower, you can retake the test, using the insights from your previous attempt.
Good luck!
We’d be happy to help.
Schedule a 15-minute introductory meeting at
or call +31 (0)88 00 67 180
Deepen your knowledge with Pegamento’s white papers.