A student uses AI to reorganize an assignment. An instructor uses it to draft a quiz. Another student submits an answer they cannot explain. If your school treats all three situations the same way, your AI policy needs more detail.
An AI policy for a training school should explain which uses are allowed, what students must disclose, what information must stay private, and how instructors will assess independent competence. Put the rules in the places where students actually work: course instructions, assignments, assessments, and orientation.
For certification-based programs, the central question is practical: can the learner demonstrate the knowledge and skills the course is meant to teach? A polished response is only one piece of that evidence.
Start with the learning task, not a list of banned tools
A general permission statement such as “AI is allowed” leaves too much unanswered. Does that include a take-home reflection? A practice quiz? A final assessment? A clinical record?
Divide activities into three categories:
- Independent work: AI assistance is not permitted for the specified task because the instructor needs to assess the student's own performance.
- Limited assistance: Certain uses are allowed, such as brainstorming or language editing, with boundaries and disclosure.
- Guided AI use: Working with AI is part of the assignment, and students must evaluate its output and explain their decisions.
These are suggested school policy categories, not universal legal definitions. Choose rules that match your program, approval requirements, and assessment design.
For example, an instructor might allow students to use AI to create a study schedule but require an unaided explanation of a safety procedure. The permission attaches to the activity, not the student's favorite app.
Include these seven decisions in your policy
1. Who and what the policy covers
Name the groups covered: students, instructors, administrators, and contractors where relevant. Explain whether it applies to personal accounts used for schoolwork as well as school-provided tools.
Appoint a policy owner. Someone must answer questions, approve exceptions, and update the rules when assignments or tools change.
2. Which tools the school has approved
Maintain an approved-tool list with a review date and permitted purposes. Evaluate data retention, access controls, account eligibility, contractual terms, and whether submitted material can be used for model training.
Approval for drafting public marketing copy does not automatically mean approval for uploading student records. The U.S. Department of Education's guidance on online educational services is a useful starting point for education-technology privacy review. Determine which privacy requirements apply to your particular school.
3. What students may submit
Explain the difference between receiving help and representing someone else's work as your own. If a student uses AI to improve grammar, must they retain the original draft? If AI helps with an outline, is a short disclosure enough?
Use examples. Students should not have to interpret an abstract paragraph while facing a submission deadline.
4. What information must stay out of AI tools
Your policy should prohibit putting patient-identifying information, passwords, payment details, confidential school documents, or identifiable student records into unapproved tools.
For class practice, use instructor-created fictional cases. Removing a name from a real story may still leave enough details to identify someone.
5. How AI assistance is disclosed
A useful disclosure describes the tool, purpose, material affected, and verification performed. It should help an instructor understand the work rather than create unnecessary paperwork.
An example students can adapt:
I used an AI assistant to suggest an outline for this reflection. I wrote the final response, checked factual statements against our course materials, and kept my planning notes. I did not enter personal or patient information.
For an assignment that expressly requires AI use, ask for the relevant prompt and an explanation of what the student accepted, corrected, or rejected.
6. How instructors verify learning
Combine written work with evidence suited to the learning objective: a brief explanation, an observed demonstration, a draft history, or a supervised assessment.
Do not treat an AI detector score as a misconduct verdict. Turnitin's own guidance warns against using its AI writing result as the sole basis for adverse action against a student.
7. How concerns and appeals are handled
Publish a consistent process. Tell the student which rule and work are in question, review available evidence, allow a response, and record the decision. Provide an appeal route appropriate to your institution.
A revised policy should also have an effective date. Students need to know which rules applied when an assignment was issued.
A starter AI policy you can adapt
The following is a drafting example for school review, not a substitute for program-specific or legal requirements:
Our school permits AI assistance only within the limits stated for each activity. When an assignment does not clearly explain permitted use, ask the instructor before using AI in work submitted for assessment. You remain responsible for the accuracy and originality of your submission and for demonstrating the required knowledge and skills. Disclose permitted AI assistance using the method given in the assignment instructions. Do not enter patient information, confidential records, account credentials, or other restricted information into an unapproved AI service. Use fictional practice material supplied or approved by your instructor. AI-generated content does not establish clinical competence or replace required supervised practice. Assessment decisions remain the responsibility of authorized staff. Concerns about unauthorized assistance will be reviewed under the school's academic integrity process. Students will have an opportunity to explain their work and use the published appeal process.
Before adopting this language, add the policy owner, effective date, approved-tool location, disclosure requirements, and applicable review procedure.
Roll it out through one realistic training session
Give staff three examples: an AI-assisted study plan, a submitted reflection with disclosed editing, and an independently assessed answer generated by AI. Ask them to apply the policy and explain any disagreement.
Then repeat a shorter version with students during orientation. Include one example of acceptable use and one of unacceptable use in each assessment's instructions.
UNESCO's guidance on generative AI in education emphasizes human agency, privacy, and educational suitability. A useful local policy translates those concerns into decisions people can make during an ordinary school day.
Frequently asked questions
Should a training school ban AI completely?
Some assessments may require a complete restriction. A school-wide policy can still distinguish independent assessment from permitted study support and instructor preparation. Make those boundaries explicit.
Does disclosure automatically make AI use acceptable?
No. Disclosure explains what happened; it does not override a task's rules. An openly disclosed use can still violate an independent-work requirement.
How often should the policy be reviewed?
Set a regular review point, such as before each term, and review earlier when a significant tool, program requirement, or incident changes the risks. Keep previous versions available for resolving questions about earlier assignments.
Put the policy alongside the learning
Steams Online's learning management tools bring course materials, assignments, quizzes, and progress into the same training environment. Use your course instructions to make expectations visible before work begins.
When reviewing your setup, book a Steams Online demonstration and walk through where students see instructions, submit work, and receive feedback. Clear rules are most useful when they are part of the actual course experience.
