Responsible Use in Education

Jason Blomquist; Liza Long; Joel Gladd

8 Responsible Use in Education

Responsible Use Applies Throughout the Workflow

Evaluating AI Outputs and Interactions taught the first three parts of ABC-R. Accuracy and Evidence checks the claim, Bias and Perspective checks the frame, and Context and Relevance checks the fit. Those checks help you judge an output.

The R asks a different question: Should this AI use happen under these conditions? It may arise before, during, or after use. Privacy and tool access often need attention before the first prompt.

An output can be accurate and useful while the use remains irresponsible. The problem may be student information that should not have been uploaded, feedback that replaces instructor judgment, or a required tool some students cannot access. Responsible Use examines those conditions.

The U.S. Department of Labor’s AI Literacy Framework describes responsible AI use through five practical concerns: sensitive information, workplace or legal requirements, misuse or harm, risks that change with the situation, and accountability for outcomes. This course turns those concerns into four questions for faculty.

Four Questions for Responsible AI Use
Question	What to examine
What must remain human?	The judgment, relationship, expertise, or learning the task requires.
What can the tool see or do?	The data, accounts, files, permissions, and actions involved.
Who else is affected?	Students, colleagues, authors, workers, communities, and people asked to trust the result.
How will the use be accountable?	The policy, approval, review, documentation, disclosure, or citation the situation requires.

These questions are this course’s practical synthesis. They remain useful as tools change because they focus on conditions of use.

Who Handles Which Responsible Use Decision?
Role	Responsibility
Faculty member	Define the purpose, protect the learning and judgment, design an accessible activity, inspect the result, and disclose meaningful AI assistance.
Technology, privacy, or security staff	Evaluate approved accounts, contracts, data handling, permissions, and security.
Accessibility and instructional support	Advise on tool accessibility, accommodations, and equivalent learning routes.
Purchasing, legal, or research-oversight staff	Address licensing, copyright, data governed by law or institutional policy, and research requirements when needed.

Office names differ. Faculty own course design and pedagogical judgment. Specialists advise on accessibility and handle institutional privacy, security, contract, and compliance processes. Approval alone does not make every use responsible.

What Must Remain Human?

Human oversight means that a person reviews the AI’s contribution, makes the consequential decisions, and accepts responsibility for the result. The person needs enough knowledge and authority to question the tool. Clicking approve does not count as meaningful oversight when the reviewer cannot explain what was checked.

The amount of oversight should match the stakes. Using AI to reorganize headings in your own planning notes requires a different level of review than using it to recommend grades, write student feedback, interpret a policy, or identify students who may need intervention.

Automated grading makes the boundary visible. Marc Watkins describes tools that promise to evaluate student work and generate feedback. A faculty member remains responsible for whether the evaluation is accurate, fair, aligned with the assignment, and appropriate for that student. In many cases, the educational relationship itself requires the instructor to read the work and make the judgment.

Learning creates another boundary. An accurate answer can still remove the thinking a student needs to practice. Ask what ability the task is meant to develop. If students are learning to interpret data, build an argument, diagnose a problem, or explain a concept, AI should not quietly perform that ability for them.

This concern applies to faculty too. AI can organize material or reduce repetitive formatting, but accepting its synthesis without doing the reading and judgment yourself can weaken your grasp of a course, policy, or student situation. In two studies of AI-supported logical reasoning, participants performed better while also overestimating their performance. Fernandes et al. (2026) describe this as a gap between performance and metacognition, or the ability to judge what you understand.

Use two checks:

Judgment check: What decision or interpretation must remain mine?
Learning check: What ability is this task supposed to strengthen, and will this use help someone practice it?

The answer may support AI use, limit its role, or rule it out for that task.

What Can the Tool See or Do?

Responsible use starts before you enter a prompt. Check the account, information, and permissions involved.

The Family Educational Rights and Privacy Act (FERPA) protects education records: records directly related to a student and maintained by an institution or a party acting for it. Grades, submissions, private feedback, and discussion posts may qualify. The Department of Education’s de-identification guidance explains that removing a name may not de-identify the material because other details can identify a student alone or in combination.

The U.S. Department of Education advises educators to check whether an online service is institutionally approved and to consult the appropriate technology or administrative office. An institution may have a contract that controls how a service handles student information. A personal account with the same company may have different terms. The logo alone does not tell you whether a workflow is approved.

Privacy and Approval Checks Before Using AI with Student Information
Situation	Responsible starting point
Names, student IDs, grades, advising details, or private feedback	Do not enter them into an unapproved tool. Follow the institution’s approved process.
Student submissions, discussion posts, recordings, or class transcripts	Check institutional guidance before uploading. The work may identify a student even after obvious names are removed.
A hypothetical teaching example	Use invented names and details that cannot be traced to an actual student.
A tool requesting access to a drive, inbox, browser, or learning management system	Check approval and permission scope. Give the least access needed, use copies, and keep sensitive accounts disconnected when possible.

Tools that can take more actions on their own deserve a stricter check. A chatbot sees what you enter. An AI agent can carry out several steps, such as reading files, inspecting browser pages, connecting to an account, editing documents, or taking actions. Before using one, identify what it can see, what it can change, and which actions require your approval. Work with copies and require confirmation before sending, submitting, deleting, purchasing, or changing the only version of a file.

The NIST Generative AI Risk Management Profile treats privacy, security, human oversight, and documentation as organizational responsibilities. Faculty should know where to ask about tool approval, problems, and sensitive information.

Use two checks:

Information check: Whose information would enter the tool, and could the person be identified?
Permission check: What account or files can the tool access, what can it do, and has this workflow been approved?

When the answer is unclear, pause the workflow and ask the office that owns the policy or system.

Who Else Is Affected?

Responsible use widens the view beyond the person entering the prompt. A stakeholder is someone affected by the decision or reasonably asked to trust the result.

Some stakeholders are easy to see. Students are directly affected when AI helps evaluate their work. Uploading a shared document without permission affects the colleague who created it. Requiring a paid AI feature affects the whole class.

Other stakeholders sit farther from the task. Authors and artists may care how their work is used. People who label training data or screen harmful content help make AI systems function. Communities may experience the costs and benefits of data-center development, while institutions choose contracts and acceptable levels of risk. NIST’s AI Risk Management Framework considers effects on people, organizations, communities, society, and the environment.

A Quebec study of 1,198 students, parents, and educators in primary and secondary education found that acceptance and trust varied by stakeholder and use. Although not a higher-education study, Karran et al. (2025) support this caution: affected people may judge the same use differently.

Faculty do not need to solve every system-level problem before using an AI tool. They do need to notice which consequences belong to the decision they are making.

Accessibility belongs in that decision. The U.S. Department of Education requires institutions to provide equal access to educational benefits delivered through digital technology. If you require an AI tool, design an accessible activity and an equivalent route to the learning outcome. An AI-enabled assistive feature may provide access without doing the learning the assignment measures. For example, reviewed AI-generated captions might help a student access a video while the student still completes the interpretation the assignment measures. Ask institutional accessibility or instructional-technology staff when the route is unclear.

Consider three examples:

Requiring a paid chatbot for an assignment creates an access question. Is there an institutionally provided option or a no-cost route that reaches the same learning outcome?
Uploading a colleague’s slides or a publisher’s test bank creates an ownership and permission question. Do you have the right to provide that material to the tool?
Adopting AI-generated feedback for a whole program creates a labor and trust question. Who reviews the feedback, who handles errors, and what work is being shifted away from faculty or staff?

Leon Furze’s Teaching AI Ethics project is a useful deeper resource for examining privacy, data, copyright, labor, and power. The purpose here is smaller: identify the people involved and the values that could be affected before deciding.

Use two checks:

Stakeholder check: Who provides the data, receives the output, relies on the decision, or bears the risk?
Access check: Does this use assume a paid tool, particular device, reliable internet, or level of technical confidence that not everyone has?

If another person’s work, data, opportunity, or trust is involved, include that person in the decision.

How Will the Use Be Accountable?

Accountability makes the human role visible. It includes following policy, reviewing the result, explaining the use, and keeping internal documentation when the stakes or policy require it.

An AI disclosure statement describes how AI contributed to a particular work product. A useful disclosure usually answers four questions:

What tool did I use?
What did I ask it to do?
What did I check, change, or reject?
What judgment and responsibility remained mine?

For course materials, a disclosure might read:

I used Microsoft Copilot to suggest a clearer order for the headings in this study guide. I selected the final organization and wrote the explanations and examples. I checked all claims against the assigned course materials and remain responsible for the final guide.

For a committee report, it might read:

I used Claude to compare the headings in three public policy documents and create an initial comparison table. I checked the table against the original documents, corrected two classifications, and wrote the analysis and recommendations.

Disclosure is different from citation. A disclosure explains the role AI played in the work. A citation points readers to material that you quoted, paraphrased, analyzed, or treated as a source. Some situations require one; some require both.

Internal documentation records enough of a higher-stakes workflow for review. Routine uses may require no saved record; grading, research, purchasing, or program-level decisions may require one under institutional policy.

Current style guidance is still developing. The MLA’s updated guidance recommends citing AI-generated material when it is quoted, paraphrased, or incorporated into a work. Its separate guidance on describing AI use explains why substantive use may require an acknowledgment beyond a citation. The Chicago Manual of Style often treats a statement in the text as sufficient, while allowing a note when a formal citation is needed.

Do not cite a chatbot as a substitute for the source it summarized. Open the original source, read it, and cite that source directly. When the AI output itself is the object you are discussing, preserve enough information for a reader to understand the interaction: tool, model if known, date, prompt or task, and a stable link or saved record when one is available.

Follow the requirements of your institution, discipline, publisher, grant, or professional setting. A disclosure statement does not override a policy that prohibits the use.

Sarah Makes a Responsible Use Decision

Sarah wants help identifying patterns across a discussion forum in her composition course. The task looks useful, and she pauses before uploading anything.

First, she asks what must remain human. Reading students’ posts and interpreting their learning are part of her responsibility as the instructor. AI may help her organize notes, but it should not replace her reading or decide what students understood.

Next, she checks what the tool can see. Her institutional Copilot account is approved for general work, but she cannot find guidance allowing her to upload identifiable student discussion posts. She does not upload the posts. She reads them herself and writes a short summary of three recurring patterns, using no names, quotations, or details that could reasonably identify a student. After confirming that her institution allows this lower-risk use, she asks Copilot to organize her notes into a comparison table.

Then she considers who is affected. Students may reasonably expect their posts to be read by the instructor and classmates named in the course directions. They may not expect their writing to be sent to another system. Sarah’s narrower workflow respects that boundary and gives every student the same instructional benefit.

Finally, she makes the use accountable. She checks the table against her notes, corrects one category, and tells the class what she did:

I read your posts and identified three recurring ideas. I used our institutionally supported Copilot account to organize my non-identifying notes into a comparison table. I checked and revised the table before sharing it. No student posts, names, or quotations were entered into the tool.

Sarah’s decision is responsible because she can explain the tool’s role, the information boundary, the people affected, and the judgment she retained.

Try This: Write a Responsible Use Decision Note

Choose one way you are considering using AI in your teaching or professional work. Use a real task, such as organizing course materials, creating practice questions, reviewing a public policy, or preparing a committee document.

Write a short note using these prompts:

Proposed use. I am considering using [tool or type of tool] to ____.

Human responsibility. The judgment, relationship, expertise, or learning that must remain human is ____.

Information and permission. The tool would see or access ____. The policy, approval, or safer boundary I need is ____.

People affected. The people who provide data, receive the output, rely on the result, or bear the risk are ____.

Accountability. I would make the use accountable by reviewing, disclosing, or citing ____.

Decision. I will use it / narrow the workflow / choose a different tool / not use AI because ____.

If your decision is to use AI, add a two- or three-sentence disclosure statement. The statement should describe what the tool did, what you checked or changed, and what responsibility remained yours.

What to Keep

Evaluating AI Outputs and Interactions evaluates the output. Responsible Use evaluates the conditions before, during, and after using AI.
Human oversight requires a person who can question the tool and remains responsible for the decision.
Check institutional approval before student information enters an AI system. Use the least data and access needed.
Identify people affected by the workflow. If you require a tool, design an accessible way to reach the same learning outcome.
Disclosure explains AI’s role. Citation documents AI-generated material used as a source. The context may require one or both.
A responsible decision can be to use AI, narrow the workflow, choose a different tool, or not use AI.

Looking Ahead

This chapter concludes Part 1. You now have a foundation for understanding AI, directing it, evaluating its outputs, and deciding whether a particular use is responsible.

Part 2 turns toward course and assignment design. It examines learning outcomes, AI-use expectations, authentic assessment, visible learning processes, and accountability. The Responsible Use questions from this chapter will return whenever you decide what role AI should play in a course or assignment.

References

Autio, C., Schwartz, R., Dunietz, J., et al. (2024). Artificial Intelligence Risk Management Framework: Generative Artificial Intelligence Profile. National Institute of Standards and Technology. DOI record

Chicago Manual of Style. (n.d.). Citation, documentation of sources: Generative AI. Source link

Fernandes, D., Villa, S., Nicholls, S., et al. (2026). AI makes you smarter but none the wiser: The disconnect between performance and metacognition. Computers in Human Behavior, 175, 108779. DOI record

Furze, L. (2026). Teaching AI ethics. Source link

Karran, A. J., Charland, P., Trempe-Martineau, J., et al. (2025). Multi-stakeholder perspective on responsible artificial intelligence and acceptability in education. npj Science of Learning, 10, 44. DOI record

Modern Language Association. (2025, August 13). Beyond citation: Describing AI use in your work. Source link

Modern Language Association. (2025, August 13). How do I cite generative AI in MLA style? Updated and revised. Source link

National Institute of Standards and Technology. (2023). Artificial Intelligence Risk Management Framework (AI RMF 1.0). DOI record

U.S. Department of Education, Student Privacy Policy Office. (2012). Data de-identification: An overview of basic terms. Source link

U.S. Department of Education, Student Privacy Policy Office. (n.d.). Family Educational Rights and Privacy Act (FERPA). Source link

U.S. Department of Education, Student Privacy Policy Office. (n.d.). I want to use an online tool or application as part of my course. However, I am worried that it is a violation of FERPA. What should I do? Source link

U.S. Department of Education, Office for Civil Rights. (n.d.). Disability discrimination: Technology accessibility. Page last reviewed January 14, 2025. Source link

U.S. Department of Labor, Employment and Training Administration. (2026, February 13). Training and Employment Notice No. 07-25: The U.S. Department of Labor’s Artificial Intelligence Literacy Framework. Source link

Watkins, M. (2025, October 10). The dangers of using AI to grade. Rhetorica. Source link

License

Icon for the Creative Commons Attribution-NonCommercial 4.0 International License

A Guide to Teaching and Learning with Artificial Intelligence Copyright © by Jason Blomquist; Liza Long; and Joel Gladd is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License, except where otherwise noted.

8 Responsible Use in Education

Responsible Use Applies Throughout the Workflow

What Must Remain Human?

What Can the Tool See or Do?

Who Else Is Affected?

How Will the Use Be Accountable?

What to Keep

Looking Ahead

References

Further Reading

License

Share This Book