AI Tools & Policy

AI-powered coding assistants can significantly accelerate your data science learning journey. However, it is critical to use them safely and in compliance with University regulations.

The “Traffic Light” System

The University of Leeds uses a traffic light system to indicate how AI can be used in assessments. Always check your specific module guidelines.

  • 🔴 Red: No AI permitted. You must do all the work yourself.
  • 🟠 Amber: AI permitted as a supportive tool (e.g., for ideation, debugging, or explaining concepts) but not for generating the final output.
  • 🟢 Green: AI is integral to the assessment and can be used freely (with appropriate acknowledgement).

1. The “Official” Route (Safe for Coursework)

Microsoft 365 Copilot (Enterprise) is the University-endorsed tool.

  • Why: This version is private. Your data is not used to train the model (see University guidance on GenAI tools). This is the recommended tool for university work where AI is permitted.
  • Access: Log in to Microsoft 365 with your University credentials (@leeds.ac.uk).
  • Usage: Use it to draft text, summarize documents, or explain code concepts securely.

2. The “Developer” Route (For Personal Skills)

GitHub Copilot is an industry-standard AI pair programmer.

  • Context: While excellent for coding, the standard/student version may not have the same data protections as the University’s Enterprise tools.
  • Usage: Use this for personal learning and building your developer portfolio.
  • Warning: ⚠️ Do not paste unpublished assessment questions, sensitive research data, or participant information into these tools.

Getting Access (Student Pack)

Students can get free access to the GitHub Student Developer Pack (a valuable industry perk):

  1. Visit GitHub Education
  2. Apply for the Student Developer Pack using your .ac.uk email (for verification only).
  3. Once verified, you can install the GitHub Copilot extension in VS Code or RStudio.

Best Practices

TipWriting Good Prompts
  • Write clear comments describing what you want to do
  • Use descriptive variable names
  • Break complex tasks into smaller steps

DOs ✅

  • Use AI to understand error messages
  • Ask AI to explain unfamiliar code
  • Use suggestions to learn new approaches
  • Verify AI-generated code before using it

DON’Ts ❌

  • Don’t blindly copy-paste without understanding
  • Don’t use AI to do all your work (you won’t learn!)
  • Don’t share sensitive/personal data with AI tools
  • Don’t assume AI code is always correct
WarningCritical Thinking Required

AI tools are powerful assistants, but they can make mistakes. Always review and test the code they generate!

Resources

Reuse