Session Teaching Card
Direct and Improve the Machine
How do I remain responsible for AI-assisted work?
Objective
Evaluate usability, accuracy, accessibility, coherence, privacy, and responsible tool use.
Method
AI critique, verification, and accessibility review
Evidence
Decision log
Setup Notes
- Prepare a critique checklist for usability, accuracy, accessibility, coherence, privacy, and ownership.
- Create a decision log template with accept, reject, revise, verify.
- Bring examples of AI output that looks plausible but needs human review.
Key Teaching Points
- Responsible AI use requires review, verification, and documentation.
- Quality improves when students critique the output against the product promise.
- A decision log makes ownership and judgment visible.
Activity Guidance
- 1Students review product v1 against the checklist.
- 2They document at least three decisions in the log.
- 3They make one targeted improvement and explain the reason for it.
Discussion Prompts
- What did the machine get right, miss, or make up?
- What is unclear, inaccessible, inaccurate, or off-purpose?
- Which decisions are yours, and how can someone tell?
Reflection
What did I accept, reject, revise, or verify?
Teacher Watch
Students may assume working means good. Push for evidence against criteria.
Support
Have students evaluate one page, one interaction, or one content section first.
Extension
Ask for an accessibility pass plus a privacy or bias risk note.