AI Policies

AI Use Policy

Overview

AI tools (e.g., ChatGPT, GitHub Copilot, Gemini, Claude) are allowed for debugging (e.g., identifying syntax errors, explaining compiler issues, clarifying error messages), testing, brainstorming, and clearly labeled bonus or extension tasks. However, AI cannot be used for core assignments unless explicitly permitted in the assignment’s grading section.

Debugging with AI does not include generating or restructuring full solutions. When using AI, you must include an AI Reflection in the README. Failure to disclose permitted AI use or using AI for core work constitutes an academic integrity violation.

Full Policy

Purpose & Principle

We use AI as a learning tool to learn faster, debug quicker, and explore more creative ideas, but not to outsource core learning.

Default Rule

• Core individual assignments: AI-generated code or writing is not allowed unless an assignment explicitly says otherwise. You may still use AI for debugging, testing, and explanation (see §3), with disclosure.
• Bonus/extension/creative portions: AI is permitted (and sometimes encouraged); you must disclose usage.

What is Allowed vs. Not Allowed

Allowed (with disclosure)

• Debugging & Testing: Explaining errors, suggesting fixes, generating tests or edge cases, test data or mocks, validating logic, and regression test suggestions.
• Study helper: Explaining errors, suggesting fixes, and generating test cases.
• Brainstorming & creativity: Feature ideas, UI/CSS variations, performance or security suggestions.
• Code quality & security: Refactoring advice, readability improvements, secure configuration reminders.
• Starter scaffolds for bonus or creative portions (where the rubric says “AI allowed”).

Not Allowed (unless explicitly permitted)

• Submitting AI-generated code or text as your core individual deliverable.
• Using AI to solve graded problems end-to-end (design → implementation → tests) without your own work.
• Submitting undisclosed AI assistance.

Examples:

• OK: “Why is Apache returning 403?” “Generate edge cases for this SQL query.” “Explain this regex.” “Suggest tests for this function.” “Refactor for readability.”
• Not OK (core, unless allowed): “Write the full solution for Assignment 2.” “Build my final project backend.”

AI Disclosure (required when AI is used for anything)

Include an AI Reflection section in each submission (README). See AI Best Practices for details.

Higher Standard when AI is used

• Explainability: You must be able to defend design decisions and trade-offs.
• Originality: You must show personal contributions (commits, comments, write-ups).
• Oral check readiness: You may be asked to walk through your code.

Integrity and Consequence

• Undisclosed or prohibited AI use is an academic integrity violation.
• Instructors may compare chat traces to submissions, run similarity and authorship checks, and review commit history.