Assessment Rubrics for AI-Integrated Learning

Generic and customizable rubric framework for assessing assignments that integrate AI tools (chatbots, LLMs, RAG systems, code assistants, creative AI, etc.) across disciplines.

© 2026 Michael Borck, AI Facilitator, School of Management and Marketing, Curtin University

For Curtin University use. For external distribution or use outside Curtin University, see LICENSE.md.

For academic staff: Start with the documents in outputs/docx/ in numerical order.

---

Quick Start

For Users (Academic Staff)

1. Download the Word documents from outputs/docx/:

   • 1_README_Getting_Started_v2.docx — Start here
   • 2_Generic_Rubric_Template_v2.docx — Your framework
   • 3_Adaptation_Guide_v2.docx — How to customise
   • 4_Example_CloudCore_Audit_v2.docx — Worked example

2. Follow the workflow outlined in the README

3. Adapt the template for your unit

For Maintainers/Developers

Prerequisites:

• pandoc (for markdown → docx conversion)
• git

Workflow:

1. Edit markdown files in sources/:

   # Make changes to any .md file
   nano sources/2_Generic_Rubric_Template.md

2. Regenerate Word outputs:

   chmod +x build.sh
   ./build.sh

3. Commit changes:

   git add sources/ outputs/
   git commit -m "Update rubric criteria and examples"

---

Project Structure

assessment-rubrics-for-ai/
├── sources/                              # Single source of truth (markdown)
│   ├── 1_README_Getting_Started.md
│   ├── 2_Generic_Rubric_Template.md
│   ├── 3_Adaptation_Guide.md
│   └── 4_Example_CloudCore_Audit.md
├── outputs/
│   ├── docx/                            # Generated Word documents
│   │   ├── 1_README_Getting_Started.docx
│   │   ├── 2_Generic_Rubric_Template.docx
│   │   ├── 3_Adaptation_Guide.docx
│   │   └── 4_Example_CloudCore_Audit.docx
│   └── html/                            # (Optional future) Web versions
├── build.sh                             # Build script (md → docx)
├── .gitignore
└── README.md                            # This file

---

Single Source of Truth

Markdown files (sources/) are authoritative. Word documents are generated from these.

Never edit .docx files directly for long-term changes. Edit the markdown, then regenerate Word.

Workflow:

Edit .md → Run build.sh → Commit both → Distribute .docx

---

Customizing Rubrics

See 3_Adaptation_Guide.md for detailed guidance on:

• Different AI modalities (chatbots, LLMs, RAG, code, creative AI)
• Discipline-specific adjustments (Business, STEM, Humanities, Law)
• Weighting by context
• Performance descriptor customization

---

Background

Developed for Curtin University's A2030 initiative by Dr Michael Borck (Business Information Systems).

Philosophy:

• Critical evaluation of AI outputs matters more than tool efficiency
• Transparency is an integrity measure, not a compliance burden
• Process (how students think) is as important as product (outcomes)
• One-size-fits-all rubrics don't work; customization is essential
• Treat AI as a "junior intern" that students must manage and validate

---

Recent Changes

v1.0 (Feb 2026): Initial release

• Generic rubric template
• CloudCore audit example
• Adaptation guide for different AI tools
• Implementation guide for staff

---

Contributing

To suggest improvements:

1. Open an issue describing the change
2. Include examples where relevant
3. Consider disciplinary context

To add new examples:

1. Create a new file in sources/ (e.g., 5_Example_Code_Assistant_Rubric.md)
2. Update numbering in README
3. Run build.sh
4. Commit and push

---

License & Terms of Use

© 2026 Michael Borck, AI Facilitator, School of Management and Marketing, Curtin University

For use within Curtin University. External distribution, reproduction, or use outside Curtin University requires explicit written permission.

See LICENSE.md for full terms.

---

Contact

Dr Michael Borck School of Management & Marketing Curtin University

---

Last updated: February 2026