Confidence Mismatch

In this worksheet, students analyze examples of “confidence mismatch,” where an AI gives an incorrect answer with absolute certainty. For each question and AI output, learners identify the mismatch and explain why the situation could be dangerous in real-life contexts such as school, research, or decision-making. Reflection questions guide students to think about the risks of trusting confident-sounding AI responses and develop strategies for verifying information. This activity builds awareness of AI limitations and encourages healthy skepticism and fact-checking habits.

Curriculum Matched Skills

English Language Arts – Critical Evaluation of Informational Text

Technology Literacy – Understanding AI Limitations

Critical Thinking – Identifying Misinformation Risks

Social Studies – Responsible Digital Citizenship

This worksheet is part of our Why Do AI Models Make Mistakes collection.