A concern is a starting point. It is not a finding of misconduct.
A fair, learning-focused response
Work through these five steps
Check what was communicated
Revisit the assignment instructions, syllabus, and any later clarification. What AI use was prohibited, permitted, or required to be disclosed? Was the distinction between generating content and editing it clear?
If expectations were unclear, acknowledge that. Consult the applicable process before deciding on a response; don’t retroactively add an AI rule. Clarify expectations for future work.
Separate observations from assumptions
Record the specific passage, source, calculation, or mismatch with the assignment that concerns you. Check references and facts yourself. A missing or inaccurate source is a concern to explore, but it does not identify how the work was produced.
A change in style, polished grammar, a formal tone, or a detector score does not establish unauthorized AI use. Avoid assumptions based on age, language background, disability, or how you expect a student to write.
On detector limitations and privacy: Vanderbilt’s guidance on AI detection.
Invite a private conversation
Explain the concern neutrally and hear the student’s account before reaching a conclusion. Give advance notice of the assignment and the points you want to discuss. Avoid public accusations or confronting a group about suspected individual misconduct.
Ask about their process, sources, and tool use. Where available, invite relevant notes, drafts, or version history; the absence of these is not proof of misconduct. Don’t demand passwords, access to personal accounts, or unrelated chat histories.
Allow time to respond and appropriate accessibility supports or another suitable way to explain. Nervousness, memory, language, and communication differences can affect an impromptu explanation.
Consider the whole account
Distinguish permitted editing or accessibility support from unauthorized generation. Ask what the tool actually did and compare that use with the instructions students received. Naming a tool alone doesn’t establish a breach.
An explanation can help you understand the student’s learning and process. A fluent explanation does not prove authorship; difficulty explaining does not prove AI use.
Grade the submitted work against the published criteria. Don’t introduce a surprise oral test or an automatic zero as a shortcut around the academic-integrity process. Any additional assessed demonstration or reassessment should follow the applicable process and accommodation requirements.
Document and choose the next step
Keep a factual record of the original instructions, the specific concern, the student’s explanation, relevant materials, and the follow-up. Separate what you observed from what you inferred, including information that resolves the concern.
Use approved institutional storage and share only with those involved in the process. Don’t upload identifiable student work to an unapproved AI tool or detector.
Where a concern remains, consult your program coordinator or academic supervisor about the applicable procedure. Explain the next steps to the student, including any response or review opportunities required by that procedure.
Words you can adapt
Open with curiosity and a specific concern
“I’d like to talk about how you developed this assignment. I noticed [specific observation], and I want to understand your process before drawing any conclusions. Could you walk me through how you approached it and what tools or support you used?”
Useful follow-up questions
- “What did you understand the AI-use expectations to mean for this assignment?”
- “How did you develop this idea, and which source helped you reach it?”
- “What changed between your first attempt and your final submission?”
- “Did any tools help you brainstorm, translate, edit, or generate content? What did you keep or change?”
- “Is there anything else you would like me to consider?”
Use these as conversation starters, not a checklist the student must pass. Close by summarizing what you heard and what will happen next.
After the conversation
Match the response to what you know
The concern is resolved
Let the student know. Record the clarification and assess the work using the usual published criteria.
Expectations were unclear
Explain the intended boundary and agree on clearer expectations going forward. Ask your coordinator which educational response is available under the existing process; don’t assume a resubmission or penalty is yours to impose.
The evidence is inconclusive
Do not treat unresolved suspicion as an established breach. Seek advice if needed, document the uncertainty, and apply the published assessment criteria without adding a suspicion-based penalty.
There is a supported concern or the student reports prohibited use
Clarify what happened and follow the applicable academic-integrity procedure. A learning-focused response can coexist with required reporting; this guide does not authorize bypassing it or selecting a particular sanction.
Make the next assignment clearer
Explain the boundary before work begins
Specify what students need to do themselves, which kinds of assistance are allowed, and what they should disclose. Give examples of the difference between feedback, editing, and generating an answer, and explain why the boundary supports the learning goal.
Build proportionate process evidence into the assignment from the outset: a short proposal, a source check, a draft, or a reflection on a revision. Make expectations and any assessment criteria explicit and accessible.
The student AI orientation can support a class discussion about permission, checking understanding, and honest disclosure. It does not set the AI rules for your course.
Continue reading or listening
Resources behind this approach
These external resources offer teaching perspectives. They are not Great Plains College policy.
- Teaching for Integrity in the Age of AI
Teaching in Higher Ed, episode 568. Tricia Bertram Gallant and David Rettinger discuss their book The Opposite of Cheating, responding respectfully, and following institutional procedures.
- Guidance on AI detection
Vanderbilt University explains its concerns about false positives, privacy, and using detection results to judge student work.
- International Center for Academic Integrity
A starting point for academic-integrity resources and professional learning.
Sources reviewed September 29, 2026.