DS307: Data Ethics & Responsible AI

by

SFIA Reference: SCAD Level 4

Data Scientist (Grade 3)

Competency Description

Ensures responsible AI practices, including fairness, explainability, and compliance with data privacy regulations.

Learning Outcomes

  • Apply bias detection and mitigation techniques.
  • Use tools for explainability (e.g., SHAP, LIME).
  • Document ethical considerations and risk.

Evidence Requirements

Level 4:

  • Used a fairness or explainability tool in a live model.
  • Identified potential bias in data or outcome. Level 5:
  • Led a model ethics review.
  • Trained others in responsible AI tooling.

Suggested Learning Activities

  • Ethics in AI internal training.
  • SHAP or Fairlearn lab walkthrough.
  • Reflection: Case study on model bias.

Recommended Resources

  • Microsoft Responsible AI Toolbox
  • KnowHow: “Fair Models – Real Stories”

Leave a Comment