All courses

Intermediate · module I10

Responsible AI in Practice

By the end: Audit a dataset and a model for bias, document a model responsibly, and explain the main AI laws and policies

Start here
Covers
Measuring and reducing bias, protecting privacy, explaining models, and the laws that apply
Contents
4 chapters, 13 topics
Time
About 2 hours
Level
Intermediate
Your progress
0 / 13

This module builds on Machine Learning Foundations. We suggest looking at it first, but the choice is yours.

Chapter 1 Ethics in Depth

  1. 1.1

    Ethical Frameworks for AI

    Rights-based, consequence-based and virtue-based thinking, and bioethics

    Case study: the extra-care list

    12 minBeing written
  2. 1.2

    The Five Pillars of AI Ethics

    Fairness, transparency, privacy, safety and accountability

    Interactive: score a product against the pillars

    10 minBeing written
  3. 1.3

    Moral Dilemmas and AI

    Trade-offs where every option has a cost

    Lab: explore moral dilemma scenarios

    10 minBeing written

Chapter 2 Fairness and Bias

  1. 2.1

    Measuring Bias in Data

    Counting and comparing groups in a dataset

    Code: audit a dataset

    12 min Activity
  2. 2.2

    Measuring Bias in Models

    Accuracy and error rates for each group

    Code: compare results across groups

    12 minBeing written
  3. 2.3

    Reducing Bias

    Better data, re-weighting and human review

    Code: improve a biased model

    12 minBeing written

Chapter 3 Transparency and Privacy

  1. 3.1

    Explainable AI

    Feature importance and explaining a single prediction

    Code: explain a model's decision

    12 minBeing written
  2. 3.2

    Model Cards and Datasheets

    Documenting what a model is for and where it fails

    Template: write a model card

    10 minBeing written
  3. 3.3

    Data Privacy and Consent

    Personal data, anonymisation and India's DPDP Act

    Case study discussion

    10 minBeing written

Chapter 4 Policy and Society

  1. 4.1

    AI Laws and Policies

    India's approach, the EU AI Act and global principles

    Interactive: classify uses by risk level

    10 minBeing written
  2. 4.2

    AI and the Future of Work

    Automation, augmentation and reskilling

    Debate

    8 minBeing written
  3. 4.3

    Environmental Impact of AI

    Energy, water and carbon costs, and how to reduce them

    Code: estimate a training run's energy

    8 minBeing written
  4. 4.4

    Module Review and Assessment

    Exercises and a scored quiz

    Graded audit report

    15 minBeing written