Trust But Verify: AI Hallucinations & Fact-Checking
Lesson 13Guided by Soleil7:00
"Soleil almost stood in front of her whole class and said something that wasn't true."
AI told her — confidently, with detail, sounding totally legit. The scariest part? AI never tells you when it's making things up. Soleil shares the exact 3-step move she uses now to catch it every time. This one might be the most important lesson in the whole program.
The big idea
AI can be confidently wrong — and your job is to catch it.
By the end of this lesson, a student can
- Explain what AI hallucinations are
- Use a 3-step fact-checking process
- Identify high-risk hallucination scenarios
Words students learn here
Also in Critical AI Use
AI can be confidently, completely wrong.
Lesson 13 is one of 16 video lessons in the AI Explorers program for grades 6–12 — each one ending in something made, not something memorized.