COMP 3270 / Module 2 / Unit 6

Unit 6 · Week 4 · Reading: Chapter 6

Human-AI Collaboration

Design collaborations in which AI roles are bounded, human authority is explicit, review is meaningful, automation bias is managed, and identifiable people remain accountable for decisions, actions, correction, and stopping.

Parent module: Prompt Engineering and Human-AI CollaborationFive lessonsEstimated study: 7–9 hoursAssignment 2 completed hereAssignment 1 continues
Unit
Unit 6 of 20
Recommended period
Week 4
Primary reading
Textbook Chapter 6
Estimated study
7–9 hours
Prerequisites
Units 4 and 5
Required platform
None
Primary framework
PARTNER
Assessment connection
Assignment 2 completed

Recommended first action: Reopen the SCOPE workflow created in Unit 5. Identify every point where an AI system assists, analyzes, recommends, drafts, modifies, or acts, then assign an accountable human reviewer and decision-maker.

Begin the role audit

Unit overview

Unit 6 examines how people and AI systems can work together without obscuring human authority or responsibility. You will classify AI roles, separate assistance from recommendation, decision, approval, and action, and identify what remains exclusively human. You will also examine automation bias, anchoring, nominal oversight, responsibility laundering, escalation, fallback, and stopping. The author-approved PARTNER framework is used to create a complete human-AI responsibility map and to finalize the workflow evidence required for Assignment 2.

Complementary roles

Assign bounded tasks based on demonstrated capability, consequence, uncertainty, and the need for human judgment.

Meaningful oversight

Ensure reviewers have relevant information, time, competence, authority, alternatives, and the ability to reject or stop.

Accountable decisions

Keep approval, consequential action, correction, remedy, and final responsibility with identifiable people or institutions.

Central idea: Human presence alone is not meaningful oversight. The human role must be informed, competent, independent enough to challenge the output, and empowered to revise, reject, escalate, or stop the workflow.

Unit learning outcomes

By the end of Unit 6, you should be able to:

  1. Distinguish collaboration, delegation, automation, augmentation, oversight, and accountability.
  2. Classify AI roles as tool, assistant, critic, simulator, recommender, or agent.
  3. Separate assistance, analysis, recommendation, decision, approval, action, correction, and accountability.
  4. Identify human-only activities requiring professional, institutional, ethical, legal, community, or relational authority.
  5. Allocate tasks among the user, AI system, specialist reviewer, decision-maker, action authority, and stop authority.
  6. Evaluate whether human review is timely, informed, competent, independent, documented, and empowered.
  7. Recognize automation bias, anchoring, overreliance, rubber-stamping, deskilling, and responsibility laundering.
  8. Design review gates that require evidence, alternatives, uncertainty, and reasons rather than simple approval.
  9. Define escalation, appeal, correction, fallback, disengagement, and stop conditions.
  10. Apply the author-approved PARTNER framework to a bounded human-AI workflow.
  11. Integrate PCAFE-V, SCOPE, and PARTNER into one coherent workflow record.
  12. Complete and self-review Assignment 2 while preserving personal authorship and required disclosure.

Unit 6 lessons

Complete the five lessons in order. These filenames are local to the dedicated Unit 6 folder.

Lesson 1 of 5 · Suggested time: 70–85 minutes

6.1 Human and AI Roles

Classify the AI role, identify complementary human strengths, and separate support activities from decisions or actions requiring human authority.

Open Lesson 1: Human and AI Roles
Lesson 2 of 5 · Suggested time: 70–85 minutes

6.2 Authority, Responsibility, and Human-Only Work

Map information, review, professional, institutional, decision, action, appeal, remedy, and stop authority.

Open Lesson 2: Authority and Responsibility
Lesson 3 of 5 · Suggested time: 70–85 minutes

6.3 Meaningful Human Oversight

Evaluate reviewer competence, information, independence, time, authority, documentation, alternatives, and capacity to reject or stop.

Open Lesson 3: Meaningful Oversight
Lesson 4 of 5 · Suggested time: 75–90 minutes

6.4 Automation Bias and Responsibility Laundering

Recognize anchoring, overreliance, rubber-stamping, capability erosion, symbolic review, and attempts to transfer responsibility to AI.

Open Lesson 4: Automation Bias
Lesson 5 of 5 · Suggested time: 90–120 minutes

6.5 PARTNER Responsibility Matrix and Module Integration

Apply the author-approved PARTNER framework, integrate PCAFE-V and SCOPE, define review and stop gates, and complete Assignment 2 preparation.

Open Lesson 5: PARTNER Integration
Folder convention: Keep unit-06.html and t-01.html through t-05.html together in a dedicated Unit 6 Moodle folder or content package.

Embedded Unit 6 study guide

This sequence concludes Module 2 during Week 4 and supports completion of the Prompt Engineering Portfolio.

  1. Classify roles
    Identify what people and AI systems do at each stage.
  2. Assign authority
    Keep review, decisions, actions, correction, and stopping accountable.
  3. Test oversight
    Audit bias, overreliance, and nominal human involvement.
  4. Integrate and submit
    Apply PARTNER and complete Assignment 2 review.

Suggested study sessions

Total: 7–9 hours
  • Session 1: Retrieval and role audit, 60–75 minutes.
    Retrieve PCAFE-V and SCOPE, read Chapter 6, and classify the AI role at each workflow stage.
  • Session 2: Lesson 1, 70–85 minutes.
    Compare human and AI contributions and identify tasks that should remain human.
  • Session 3: Lesson 2, 70–85 minutes.
    Create an authority map covering information, review, decision, action, correction, appeal, remedy, and stopping.
  • Session 4: Lesson 3, 70–85 minutes.
    Audit whether each human review gate is informed, resourced, competent, and empowered.
  • Session 5: Lesson 4, 75–90 minutes.
    Analyze automation bias and responsibility laundering, then add countermeasures and independent checks.
  • Session 6: Lesson 5 and Assignment 2 completion, 90–120 minutes.
    Apply PARTNER, integrate the Module 2 frameworks, complete portfolio self-review, and update Assignment 1.

Evidence to retain

  • AI-role classification
  • Human-only activity list
  • Authority and responsibility map
  • Meaningful-oversight audit
  • Automation-bias scenario analysis
  • PARTNER responsibility matrix
  • Integrated PCAFE-V, SCOPE, and PARTNER workflow
  • Assignment 2 final self-review
  • Assignment 1 reflection notes
Show the responsibility-allocation sequence
  1. Define the legitimate purpose and expected outcome.
  2. Identify each workflow stage and AI contribution.
  3. Classify the AI role and its limits.
  4. Identify human-only work and required authority.
  5. Assign reviewers, decision-makers, action authority, and stop authority.
  6. Provide reviewers with evidence, uncertainty, alternatives, and sufficient time.
  7. Test for automation bias and conflicts of interest.
  8. Add escalation, appeal, correction, remedy, fallback, and stopping.
  9. Document decisions and retain accountable human approval.
  10. Reassess the workflow when purpose, information, system, scale, or consequence changes.

Assessment milestones in Unit 6

Assignment 2: Prompt Engineering Portfolio

Weight: 15%

Started: Unit 4, Week 3

Final preparation: Unit 6

Recommended submission: End of Module 2, Week 4

Official Moodle due date and time: ____________________

Assignment 1: Journal continues

Reflect on human-AI complementarity, authority, meaningful oversight, automation bias, responsibility laundering, capability preservation, and changed boundaries.

Recommended final submission: End of Module 6, Week 11

Official Moodle due date and time: ____________________

Final Assignment 2 check: Verify that the portfolio demonstrates development, not only successful final outputs. Include prompt design, context and source decisions, workflow stages, role allocation, human review, verification, corrections, and required disclosure.
Show the Assignment 2 final-review checklist
  • Purpose, task, audience, format, criteria, and verification are explicit.
  • Initial and revised prompts and outputs are preserved.
  • Context is necessary, minimized, and authorized.
  • Sources and evidence boundaries are documented.
  • The task is decomposed into reviewable stages where appropriate.
  • AI and human roles are clearly distinguished.
  • Meaningful review, approval, escalation, fallback, and stopping are operational.
  • Claims and sources are verified as required.
  • Corrections and final human decisions are explained.
  • AI use is documented and disclosed according to assignment instructions.

Required learning activities

Unit 6 activities and evidence
Activity Purpose Expected product Estimated time
AI-role classification Distinguish tool, assistant, critic, simulator, recommender, and agent roles. Role map 30–40 minutes
Human-only work analysis Identify activities requiring human judgment, relationship, authority, or accountability. Boundary statement 30–45 minutes
Responsibility matrix Allocate tasks to the user, AI, specialist reviewer, decision-maker, and stop authority. PARTNER matrix 40–55 minutes
Meaningful-review audit Assess reviewer competence, evidence, time, authority, alternatives, and independence. Oversight audit 35–50 minutes
Automation-bias scenario Identify anchoring, overreliance, rubber-stamping, and responsibility laundering. Risk and safeguard record 35–50 minutes
Escalation and stop design Define thresholds, authority, fallback, appeal, correction, and remedy. Escalation pathway 35–50 minutes
Module 2 integration Connect PCAFE-V, SCOPE, and PARTNER in one accountable workflow. Integrated workflow 45–60 minutes

Participation options

Option A: Approved AI system

Use an approved system with synthetic, public, supplied, or authorized information and no consequential external action.

Option B: Supplied workflow records

Use supplied prompts, outputs, recommendations, review records, decisions, and incident scenarios.

Option C: Non-AI role analysis

Complete the same responsibility, oversight, bias, escalation, and workflow analysis without operating an AI platform.

All options assess the same collaboration and accountability outcomes.

Responsible-use boundaries and support

Boundaries for Unit 6

  • Do not upload confidential, personal, proprietary, privileged, security-sensitive, or culturally governed information.
  • Do not permit AI to approve, decide, publish, send, transact, delete, or take consequential action.
  • Do not describe nominal human presence as meaningful oversight.
  • Do not transfer professional, institutional, legal, ethical, or community responsibility to AI.
  • Stop when a qualified reviewer, adequate evidence, correction pathway, or accountable authority is unavailable.

Support and accessibility

  • Use supplied records if live AI is inaccessible, unavailable, unsuitable, or unwanted.
  • Provide a text description for every visual responsibility or workflow map.
  • Contact the course coordinator for assignment, authority, or permitted-use questions.
  • Seek specialist, accessibility, privacy, security, ethics, or professional guidance when appropriate.

Unit 6 and Module 2 completion checklist

Unit 6 completion items

Previous: Unit 5Return to Module 2Continue to Module 3: Responsible and Ethical AI Use
👤