Strategies for teaching responsible AI use

Start with the learning goal. Identify the reasoning, practice, or judgment students must demonstrate themselves.

Ask one question: What would students fail to learn if AI did this part for them? Use the answer to draw the boundary between student-owned work and acceptable assistance.

Then put three statements directly in the assignment directions: 

• You may use AI to… 
• You may not use AI to… 
• Here are the reasons why… 

Place this guidance inside the assignment so students see it when they need it. 

Give students an AI response and ask them to decide what they can trust. Use a response tied to a current course topic.

Have students use this five-step check:

• Review: Read an AI response to a course-related question.
• Identify: Mark its factual claims, sources, assumptions, and perspectives.
• Notice gaps: Identify missing context, perspectives, or alternative explanations.
• Verify: Check important claims and citations against course readings or credible sources.
• Judge: Label what is accurate, questionable, incomplete, biased, or unusable.

Have students submit an annotated response or a short explanation of what they would keep, revise, or reject. 

Use a short scenario to have students decide whether and how to use AI.

Include the task, learning goal, information involved, and relevant course or professional expectations. Use it in a discussion, small-group activity, or low-stakes quiz.

Have students compare their answers and explain how the task, context, and consequences affected their judgment.

Provide the scenario and any AI output yourself so students do not need a paid account or need to share data with an external platform.

Require students to explain where and how AI affected their work. Encourage them to use disclosures that match the amount they used AI:

Use the disclosure to assess judgment: Was the purpose appropriate? Did the student verify the output? Did they remain responsible for the final choices?

Give feedback on one good decision the student made and one decision to improve next time.

Teach responsible AI use in three passes, reducing support each time:

• Model: Describe how you check an AI output, verify a claim, and decide what to keep.
• Guide: Give students clear boundaries for an assignment and provide feedback on their use of AI.
• Release: In a later assignment, ask students to decide whether AI belongs in the task and justify its role.

As students move through the three passes, shift your feedback from showing them what to do to asking them to explain their choices. This should help them progress from following guidance to exercising independent judgment.