Discipline-specific
Ground the strategy in relevant purposes, standards, evidence, professional duties, communities, and authority.
Capstone overview
The Capstone Project is the culminating assessment in COMP 327. Select a discipline-specific stream and develop a Responsible AI Integration Strategy for a bounded use case. Define a legitimate purpose, compare alternatives, identify affected people, establish information and authority boundaries, select relevant frameworks, design an accountable workflow, verify consequential claims, classify risk, propose governance and safeguards, examine future scenarios, and make a qualified recommendation.
Human-governed
Preserve review, decisions, correction, appeal, remedy, capability, participation, alternatives, and stop authority.
Evidence-supported
Verify claims and sources, document uncertainty, test scenarios, and connect recommendations to evidence.
Capstone principle: A justified conclusion may support limited
adoption, phased testing, redesign, deferral, a non-AI alternative, or stopping.
Capstone learning outcomes
- Define a bounded discipline-specific problem, purpose, intended outcome, and evidence of need.
- Compare AI-supported, human-led, professional, community-led, and non-AI alternatives.
- Identify stakeholders, affected non-users, beneficiaries, risk bearers, practitioners, and communities.
- Define the AI role and identify human-only work and authority.
- Establish information, source, permission, rights, privacy, security, and retention boundaries.
- Select and justify primary and supporting course frameworks.
- Design current and proposed workflows with responsibility and review gates.
- Verify consequential claims, sources, references, quotations, assumptions, and independence.
- Evaluate fairness, accessibility, authorship, integrity, service, capability, and relationships.
- Classify inherent and residual risk and propose proportionate controls.
- Design validation, monitoring, incidents, contestability, correction, remedy, fallback, and stopping.
- Construct future scenarios and adaptive safeguards, triggers, revalidation, rollback, and retirement.
- Integrate human purpose, agency, authority, capability, relationships, participation, and accountability.
- Present an accessible, properly cited, student-authored strategy with a qualified recommendation.
Choose a capstone stream
Select the stream that best matches your program, background, interests, community context, or career goals. The Open Interdisciplinary Stream is subject to approval.
Technology and organizations
- Computing and Information Systems
- Business and Management
- Engineering and Engineering Management
- Leadership and Entrepreneurship
- Accounting, Finance, and Economics
People, knowledge, and communication
- Education
- Health Disciplines
- Psychology, Counselling, and Human Services
- Arts, Humanities, and Social Sciences
- Communications, Media, and Journalism
- Science and Research
Public, legal, community, and environment
- Government and Public Administration
- Law, Justice, and Legal Studies
- Environmental and Sustainability Studies
- Indigenous Governance and Community Leadership
- Open Interdisciplinary Stream, subject to approval
Scope before breadth: Select one bounded use case. Avoid
topics as broad as “AI in education,” “AI in healthcare,” or “AI in business.”
Capstone development lessons
Complete the five project studios in order. Each lesson produces a concrete component of the final strategy.
Capstone Scope, Purpose, and Evidence of Need
Confirm the stream, bound the problem, establish the purpose, compare alternatives, identify evidence of need, and check feasibility.
Stakeholders, Workflow, and Responsibility Design
Map affected people, workflows, the AI role, human-only work, authority, escalation, and fallback.
Framework Selection and Integrated Analysis
Select relevant frameworks, build an evidence map, organize the analysis, and identify residual uncertainty.
Governance, Scenarios, and Adaptive Safeguards
Design risk controls, governance, validation, monitoring, incidents, remedy, scenarios, adaptive safeguards, and stopping.
Capstone Synthesis, Audit, and Submission Readiness
Audit evidence, citations, consistency, accessibility, HUMAN integration, authorship, disclosure, and submission requirements.
Folder convention: Keep
capstone-project.html and
t-01.html through t-05.html in the same Moodle folder.Two-week capstone schedule
- Early Week 12
Complete Lesson 1 and finalize scope, purpose, alternatives, and sources. - Late Week 12
Complete Lessons 2 and 3, then build the workflow, responsibility map, framework rationale, and draft. - Early Week 13
Complete Lesson 4 and finish governance, risk, scenarios, and safeguards. - Late Week 13
Complete Lesson 5, revise, verify, disclose, and submit.
Show the evidence-retention set
- Project brief, scope exclusions, evidence of need, and alternatives comparison
- Stakeholder map, workflows, AI-role statement, and PARTNER matrix
- Framework-selection rationale and claim-to-source evidence map
- Risk, control, validation, monitoring, incident, remedy, and stop records
- Four scenarios and adaptive-safeguard register
- Final HUMAN review, citation audit, AI-use record, and disclosure
Project requirements and suggested report structure
The required product is a discipline-specific Responsible AI Integration Strategy of approximately 4,000 to 5,000 words. Follow the official assignment instructions if they prescribe a structure.
Foundation and analysis
- Title page and executive summary
- Introduction, discipline, and context
- Problem and evidence of need
- Stakeholders and affected non-users
- AI and non-AI alternatives
- AI role and human-only work
- Current and proposed workflow
- Information, sources, permissions, rights, and authority
- Framework selection and integrated analysis
Strategy and recommendation
- Verification and validation
- Risk classification and safeguards
- Governance, ownership, and approval
- Monitoring, incidents, contestability, and remedy
- Future scenarios and adaptive safeguards
- Human capability, relationships, participation, and alternatives
- Residual risk and uncertainty
- Final recommendation
- AI-use disclosure, references, and permitted appendices
Framework selection
| Purpose | Frameworks |
|---|---|
| Foundational framing | AI Paradox and ORBIT |
| Prompting and workflow | PCAFE-V, SCOPE, and PARTNER |
| Rights and information | FAIRER, PROTECT, and CREATE |
| Evidence and learning | TRACE, CLAIM, LEARN, and RIGOUR |
| Applied practice | VOICE, WORKS, and LIFE |
| Discipline and integration | DOMAIN, GOVERN, FUTURES, and HUMAN |
Selection rule: Do not apply every framework mechanically. Select
those material to the use case and explain why each was chosen.
Readings and authoritative evidence
Revisit Chapters 17 through 20 and the applied chapters relevant to your stream. Use current scholarly research, professional standards, technical documentation, policies, accessibility guidance, privacy and security guidance, human-rights materials, and applicable governance resources.
Currentness rule: Verify publication date, jurisdiction,
scope, status, and continuing applicability. A draft rule, vendor claim, announcement, or old policy is
not automatically applicable.
Participation pathways
Option A: Approved AI tools
Use approved tools for bounded support, then verify all important claims independently.
Option B: Supplied records
Use supplied cases, sources, workflows, outputs, inventories, risks, incidents, and scenarios.
Option C: Non-AI development
Use course readings, databases, conventional search, word processing, diagrams, and permitted human feedback.
Live AI use, employment, organizational access, confidential records, paid databases, agents, and connected tools are not required.
Responsible use, authorship, and accessibility
Boundaries
- Do not upload confidential, personal, proprietary, privileged, security-sensitive, unpublished, Indigenous, or culturally governed information without authority.
- Do not conduct live deployment or test attacks, surveillance, transactions, deletion, or external actions.
- Do not permit AI to make consequential decisions.
- Do not fabricate sources, stakeholder views, validation results, or approvals.
- Do not claim authority you do not possess.
Authorship and accessibility
- You remain the author and accountable decision-maker.
- Verify claims, citations, quotations, references, and calculations.
- Document and disclose material AI assistance as required.
- Use accessible headings, tables, links, contrast, and text alternatives.
- Provide structured text alternatives for all diagrams.
Assignment 6 submission
Submission details
Assessment: Assignment 6, Capstone Project
Weight: 20%
Product: Discipline-Specific Responsible AI Integration Strategy
Recommended length: Approximately 4,000 to 5,000 words
Recommended submission: End of Week 13
Official Moodle due date and time: ____________________
Before submitting
- Check the assignment instructions and rubric.
- Verify claims, citations, references, and consistency.
- Confirm recommendations follow from evidence.
- State residual uncertainty and risk.
- Complete accessibility and disclosure checks.
- Confirm file format, filename, and successful upload.
Replace the placeholder anchors with final Moodle URLs.