Imagine that an AI tool tells you, “The first computer was invented in Japan in 1950.” The sentence sounds clear and confident. However, confidence does not mean truth. AI can produce false information that looks believable. This kind of error is often called a “hallucination.” (oecd.org)
Schools are becoming important places to learn how to find these mistakes. In June 2026, the OECD and the European Commission introduced an AI literacy framework for primary and secondary education. It includes 19 skills and practical classroom examples. One key goal is to help students judge the accuracy, value, and possible bias of AI-generated information. (oecd.org)
Students can practice a simple checking process. First, they should mark facts that can be checked, such as names, dates, places, and numbers. Next, they can compare the AI’s answer with a textbook, an official website, or two reliable news sources. If the sources disagree, students should examine who wrote them, when they were published, and what evidence they provide.
Students should also check references carefully. An AI tool may give the name of a book, expert, or research paper that does not exist. Instead of trusting a link or title immediately, students should search for the original source. Asking a teacher or librarian can also be helpful.
Another useful classroom activity is to give the same question to different AI tools or use different prompts. Students can compare the answers and discuss why they are different. However, agreement between AI tools is not proof. They may repeat the same incorrect information.
The final step is to explain the mistake in your own words. This is more valuable than simply finding the correct answer. When students describe what was wrong, how they checked it, and which source they trusted, they develop critical thinking. AI literacy is not only the ability to use AI. It is the habit of stopping, questioning, and checking before believing. (oecd.org)










