STEP 2: TEACH IT

September 25, 2026

This page includes information about: 

  1. Talk About AI Expectations with Your Students
  2. Teach Students How to Disclose AI Use
  3. Show Your Students, Don’t Just Tell Them 
  4. Connect AI Use to Disciplinary Practice
  5. Design For Evidence of Learning  

1. Talk about AI expectations with your students.

Discuss your AI guidance early in the semester, and revisit it before assignments.

Students should know:

  • what assistance is acceptable;
  • what crosses the line;
  • what to do when they are uncertain;
  • how AI assistance should be acknowledged; and
  • why these expectations matter in your discipline.

Simple message to reinforce with your students: When a student is in doubt they should ask before using the tool.

▶️ Download our customizable presentation deck on “AI Use Expectations”. It can be used to clarify your course priorities and present to your students about AI use expectations in your course. 

2. Teach students how to disclose AI use.

If AI use is permitted, decide what transparency should look like.

Ask your students to identify:

  • the AI tool used;
  • what they used it for;
  • how substantially it contributed;
  • which portions of the work were affected; and
  • what the student independently reviewed, revised, or verified.

The URI CYPHER Center’s A0–A5 AI Usage Marking system offers one way for students and instructors to communicate how AI contributed to an assignment. The marking system is intended to promote transparency—not to detect AI use or determine misconduct.

▶️Learn more at CYPHER AI Usage Marking Guide.

3. Show your students, don’t just tell them.

Provide examples from your discipline. General examples are provided for consideration below.

Potential Permitted Use Examples:

“I used ChatGPT to brainstorm possible research questions. From the output I received, I selected one, and based on my class notes I further developed the final question independently.”

“I used Copilot to identify possible errors in my code. I evaluated the suggestions and made the final changes myself.”

Potential Not Permitted Example:

“I generated the response with Claude and submitted the output as my own work, without reviewing it, when independent authorship was required.”

4. Connect AI use to disciplinary practice.

When appropriate, introduce the standards students may encounter in your field.

Possible resources include:

AI literacy should include more than citation. Discuss authorship, attribution, verification, accuracy, responsibility, and professional ethics within your discipline. 

Consider This: Pose this question to your students, “If you used AI this way in your future profession, what standards or responsibilities would apply?” 

5. Design for evidence of learning.

Consider assessments that make student thinking and process more visible. Also, see the AI and Assessment guidance project for more information. 

  • drafts and checkpoints;
  • brief reflections;
  • annotated sources;
  • problem-solving explanations;
  • code walkthroughs;
  • oral follow-up questions;
  • version history where appropriate;
  • course-specific applications, such as in-person quizzes and exams.

These practices support learning first. They are not intended to be assigned as surveillance.

Faculty Quick Check:

✅I talk about my AI expectations, rather than relying on the syllabus alone.
✅Students know how to disclose permitted AI use.
✅I provide examples when the boundaries may be unclear.
✅I connect expectations to disciplinary practice where appropriate.
✅My assessments give students opportunities to demonstrate their own learning.

Bottom line: Don’t just publish the rule. Teach and explain to students how to follow it.

⃪ academic integrity & AI
STEP 1: write IT
STEP 3: Address IT