AI Ethics & Responsibility

Use AI Without Handing It the Decision.

You do not need a long ethics checklist before every prompt. In OER work, a few practical boundaries cover most situations.

AI can assist. Responsibility stays with people.

Four Questions Cover Most Situations

Is this safe to share?Do not put identifying student information, confidential records, or other protected information into an AI tool unless that use is explicitly approved.
Can I verify what matters?If a claim could affect students, access, licensing, curriculum, or policy, know where you will check it.
What might be missing?Look at authorship, examples, framing, and whose perspectives or contexts may be absent.
Who owns the decision?Keep curricular, adoption, accessibility, licensing, and local policy decisions with the people responsible for them.

You do not need to solve every ethical issue before using AI. Use these questions to know when AI is useful and when the work needs to move back to a person or original source.

Keep these responsibilities human

What Responsible Use Looks Like in Practice

Protect people

Use course-level or aggregate information when possible. Do not paste student names, grades, attendance, accommodation records, personnel information, or other sensitive records into an AI system unless your institution has approved that use.

Check consequential claims

Open the original or authoritative source for important titles, links, licenses, access conditions, accessibility information, and current requirements. Confidence is not evidence.

Notice representation and framing

AI can reproduce gaps and assumptions in the material it draws from. Review who created the resources, how people and ideas are framed, and whether important perspectives or contexts are missing.

Keep accountability with people

AI can organize evidence, draft questions, and surface tradeoffs. It should not become the reason for a curricular, adoption, licensing, accessibility, or policy decision.

Two common OER traps

Free Access and AI Confidence Can Both Be Misleading

Free to read is not automatically OER

Verify reuse and adaptation rights at the original source. A public webpage, library resource, or freely available file does not automatically carry an open license.

“The AI found it” is not verification

If the tool cannot actually access the current source page, treat its claims about the resource as unverified. Ask for the source, open it yourself, and check the details that matter.

Where AI helps and where it stops

Use AI for the Setup. Move Out of the Chat for the Decision.

Good uses for AI

Generate better search terms.
Organize candidate resources.
Compare evidence you already have.
Surface unknowns and questions to verify.
Draft a librarian request or planning checklist.

Move to a person or source when

You need to confirm a license or reuse right.
You need to know what your library actually provides.
You need an accessibility determination.
You need to interpret current local policy or curriculum requirements.
You are making the final adoption or teaching decision.

Field Rule: You Should Be Able to Explain the Decision Without Saying “The AI Said So”

If an AI-assisted result affects what students read, what faculty adopt, how a course is structured, or how institutional requirements are understood, you should be able to point to the evidence, the human judgment, and the source behind the decision.

Next step

Need the Practical How-To?

The Using AI guide explains how to copy Promptbook prompts, verify results, troubleshoot weak answers, and understand the Quick, Go Deeper, Human Check, and Needs verification labels.

Choose a Task Using AI Guide Privacy