Skip to content

AI Collaboration Policy

In this course, Generative AI is not just a topic of study; it is an integral tool for learning and building. You are not only permitted but actively encouraged to use AI as a collaborative partner. Learning to leverage these tools effectively and ethically is a core learning objective.

However, the goal of the course is for you to learn. Using a tool to assist your work is different from having the tool do your work for you. This policy outlines the boundaries to ensure you are learning effectively while using AI responsibly.

✅ Green Light: Always Permitted & Encouraged

Think of the AI as your personal tutor, brainstorming partner, or pair programmer. You can and should use AI for the following tasks without any special attribution:

  • Brainstorming & Idea Generation: Asking an AI for project ideas, potential features, or ways to approach a problem.
  • Conceptual Understanding: Asking for explanations of technical concepts, definitions, or analogies (e.g., "Explain Retrieval-Augmented Generation like I'm a business manager").
  • Debugging Assistance: Pasting your code and asking an AI to help you find errors or suggest fixes.
  • Improving Your Writing: Using AI to check your grammar, fix spelling, or rephrase a sentence for clarity in your reports and documentation.
  • Code Generation with AI Assistants: Using tools like CodeAssist or GitHub Copilot to generate code snippets, complete lines of code, or create boilerplate functions is an expected part of the workflow.

⚠️ Yellow Light: Permitted with Attribution

When an AI provides a substantial, unique contribution to your submitted work, you must give it credit. This is not about penalizing you; it's about practicing professional and academic honesty.

You must provide attribution when:

  • An AI generates a significant block of text (e.g., multiple paragraphs in your README.md or project report) that you include in your work.
  • An AI provides a specific, complex algorithm or code block that is central to your project's functionality and was not generated via an integrated tool like CodeAssist.

How to Attribute:

In your group's GitHub repository, create a file named AI_CITATIONS.md. For each use case that requires attribution, add a simple entry:

Citing AI Use

Source: [Name of AI Model, e.g., Google Gemini]
Prompt: ["Your original prompt here"]
Usage: [Briefly describe how the output was used in your project, e.g., "Used as the foundational text for the 'Project Goals' section of our README.md, which we then edited and refined."]

🚫 Red Light: Academic Misconduct

Crossing the line from "use" to "misuse" happens when you outsource the core thinking and learning process to an AI. The following are considered violations of academic integrity:

  • Submitting an entire project, lab, or written assignment that was generated by an AI with little to no personal modification, input, or analysis.
  • Presenting AI-generated ideas, analysis, or code as your own original thought without attribution.
  • Using an AI to complete an assignment that is explicitly designated as an individual, tool-free assessment of your knowledge.

The Prime Directive

Here's a simple rule to follow: You must be able to explain any part of your submitted work. If you copy-paste code from an AI and cannot explain how it works, what it does, or why it's necessary, you have not used the tool correctly. The AI is your collaborator, but you are the project lead. You are ultimately responsible for and the expert on your own work.