Digital, AI & Creative Skills
Verify Before You Trust
This lesson helps students verify information before trusting or using it. They practise checking claims across sources.
What you will learn
- Understand that AI chatbots produce hallucinations because of how they are built, and that a confident tone is not a reliable indicator of accuracy.
- Learn a fast verification method that can check a specific AI claim against a real source in under two minutes, so that checking becomes a realistic habit rather than a theoretical ideal.
- Understand that not every AI output needs verification, and learn to tell the difference between claims that need to be checked and content that is not factual enough to need it.
- Combine the hallucination understanding, the fast verification method, and the needs-checking categories into a one page Verify Checklist for a real AI output the student plans to use.
What happens in this lesson
Chapter 1: Why AI Confidently Lies
Understand that AI chatbots produce hallucinations because of how they are built, and that a confident tone is not a reliable indicator of accuracy.
What you make: A saved hallucination note describing one time the student (or someone they know) encountered a confident AI wrong answer, and explaining why it probably happened.
Chapter 2: Fast Verification
Learn a fast verification method that can check a specific AI claim against a real source in under two minutes, so that checking becomes a realistic habit rather than a theoretical ideal.
What you make: A saved verification log showing three specific AI claims and the results of checking each one against a real source.
Chapter 3: What Does Not Need Checking
Understand that not every AI output needs verification, and learn to tell the difference between claims that need to be checked and content that is not factual enough to need it.
What you make: A saved needs-checking note classifying four AI output examples into 'needs verification' or 'does not need verification' and explaining why.
Chapter 4: Verify Checklist
Combine the hallucination understanding, the fast verification method, and the needs-checking categories into a one page Verify Checklist for a real AI output the student plans to use.
What you make: A saved Verify Checklist for one real AI output, identifying the claims that need checking, running the fast verification on each one, and producing a short honest summary of what was confirmed and what was corrected.
Chapter 5: When Verification Is Not Enough
Evaluate whether a verified AI output is safe, complete, and fit for purpose in a higher-stakes context, and decide what to do when it is not.
What you make: A saved trust decision note that classifies three AI outputs as use, revise, or reject, with a short justification for each decision.
Ready to learn more?
Join to access the full lesson and track progress.