ReHuman

Formation Before Permission · Decision Map

How to use the Decision Map

Use the handout to test one real AI use at a time. The goal is not to decide whether AI is good or bad. The goal is to decide what this particular use does to human learning, judgment, responsibility, and relationship—and what your institution should do next.

Before you begin

Pick one assignment, activity, policy, program, or workflow and name one specific thing AI will do. If AI is doing several materially different things, run a separate Decision Map for each use.

Work through the six steps in order.

Write short answers. You are trying to make the decision visible—not write a report.

01

Start with the purpose of the work.

Ask: “What should someone become more capable of because this work is done well?”

Name the human capacity before discussing AI. Examples: reasoning, evidence judgment, clinical judgment, disciplinary interpretation, communication, ethical responsibility, mentoring, or intellectual ownership.

02

Name the human exercise on both sides.

Student/Learner: What must the learner actually learn, practice, or become capable of doing?

Faculty/Decision-maker: What judgment, expertise, responsibility, or relationship must the human still exercise?

Then run the Formative Test: if AI removed this human exercise, would the work lose something it is intended to develop, evidence, or require? Yes = formative. No = potentially non-formative.

03

Name one specific AI use.

Ask: “What will AI actually do?”

Avoid answers like “use ChatGPT” or “use AI.” Instead write the action: “AI generates counterarguments,” “AI drafts feedback,” “AI grades the first pass,” “AI summarizes advising notes,” or “AI creates the first draft.”

04

Judge what happens to each human exercise.

Mark the student effect and the faculty effect separately. Do not force them to match.

Strengthen: the human still performs the essential exercise and AI improves support, feedback, range, or efficiency.

Bypass: a meaningful human step disappears even though the final product may still look acceptable.

Replace: AI performs the essential human work in the person's place.

05

Map both verdicts into an action.

Record a decision for the student side and faculty side. Then specify the operating conditions: AI May, Human Must, Redesign / Release Action, and Evidence / Rationale.

Decision Bridge

Strengthen Preserve / Permit
Bypass Redesign
Replace + Formative Preserve Human
Replace + Non-formative Release Task
Split verdict rule: keep both results. If one shared workflow affects both sides, the more protective condition governs implementation. If the uses can be separated, implement separate actions.
06

Run the Presence and Governance gates before implementation.

Presence Gate: ask whether care, formation, adjudication, witness, commitment/covenant, or another context makes accountable human encounter part of the value. If yes, identify what human encounter must remain.

Governance Gate: review privacy/data, approved tools/security, accessibility, intellectual property, accreditation, academic integrity, and relevant institutional policy.

A formation recommendation is not cleared for implementation while a required governance item remains “Needs Review.”

If you get Preserve / Permit

State what AI may do, what humans must still do, and what evidence will make the human exercise visible.

If you get Redesign

Identify the step AI is bypassing and change the workflow so the essential human exercise returns.

If you get Preserve Human

Do not delegate the formative human exercise. AI may still support surrounding non-essential work.

If you get Release Task

The task may be automated or removed if it is genuinely non-formative and governance requirements are satisfied.

What should you do with the result?

Treat the worksheet as a provisional decision record. Use it to improve the assignment, policy, or workflow; clarify permitted AI use; identify evidence requirements; and surface governance questions that need an owner. If the case is consequential or institution-wide, bring the result into a faculty, program, governance, or leadership conversation.

en_USEnglish
Scroll to Top