Wirewalk Instituteof AI
gov-250 Intermediate Waitlist

Responsible AI: Fairness, Bias and Explainability

For data scientists, model reviewers and responsible AI leads who must measure bias, choose and defend a fairness definition, mitigate disparities and explain model decisions to regulators and affected people.

The curriculum is being prepared and reviewed. Enrollment is not open yet.

About this course

Fairness and explainability are where responsible AI stops being a values statement and becomes engineering and evidence. This course teaches you to find bias, quantify it with the right metric for the decision at hand, decide what to do about it, and explain individual and overall model behavior in terms a regulator or an affected person can use.

You will trace bias through the lifecycle, compute and compare demographic parity, equalized odds and calibration, connect them to disparate treatment, disparate impact and the four-fifths rule, and run assessments with the open-source Fairlearn and AI Fairness 360 toolkits. You will handle missing protected attributes, test LLM applications with counterfactual prompts and benchmark suites, apply pre-, in- and post-processing mitigations, and document a search for less discriminatory alternatives. On explainability you will compare SHAP, LIME, partial dependence and counterfactual explanations, generate adverse action reasons for credit decisions, and learn the limits of LLM self-explanation. Standards covered include NIST SP 1270, ISO/IEC TR 24027 and ISO/IEC TS 6254:2025. Worked examples are hypothetical. The Wirewalk Institute certificate attests to completion and assessment of this course only.

What you will be able to do

  • Trace the sources of bias across data collection, labeling, modeling and deployment for a given system
  • Compute and interpret group fairness metrics and justify which definition fits a specific decision
  • Run a fairness assessment with Fairlearn or AI Fairness 360 and design bias tests for an LLM application
  • Select and evaluate bias mitigation techniques and document a less-discriminatory-alternatives search
  • Produce global and local explanations with SHAP, LIME and counterfactuals and state their limitations
  • Generate defensible adverse action reasons for credit or employment decisions
  • Prepare fairness and explainability documentation aligned with NIST SP 1270, ISO/IEC TR 24027 and ISO/IEC TS 6254

Syllabus

6 modules · 19 lessons · about 14 hours

Module 1: What Fairness Means and Why Definitions Collide

Sources of bias, the main group fairness metrics, and their legal anchors.
3 lessons
  1. Where Bias Enters: Historical, Representation, Measurement, Aggregation and Deployment Bias
  2. Demographic Parity, Equalized Odds and Calibration: Why You Cannot Satisfy Them All at Once
  3. Legal Anchors: Disparate Treatment, Disparate Impact and the Four-Fifths Rule

Module 2: Measuring Bias

Practical fairness assessment for classical models and LLM applications.
3 lessons
  1. Fairness Assessment with Fairlearn MetricFrame and AI Fairness 360
  2. When Protected Attributes Are Missing: Proxies, Inference Methods and Privacy Constraints
  3. Bias Testing for LLM Applications: Counterfactual Prompts, Benchmark Suites and Use-Case Evaluations

Module 3: Mitigating Bias

Intervention points, trade-offs and documentation.
3 lessons
  1. Pre-processing, In-processing and Post-processing Mitigations Compared
  2. Accuracy-Fairness Trade-offs and Searching for Less Discriminatory Alternatives
  3. Mitigation for Generative Systems: Data Curation, System Prompts and Output Filters

Module 4: Explainability Methods

Global and local explanation techniques and when each is misleading.
3 lessons
  1. Global Explanations: Feature Importance, Partial Dependence and Surrogate Models
  2. Local Explanations: SHAP, LIME and Counterfactual Explanations
  3. Explaining LLM Output: Citations, Attribution and the Limits of Self-Explanation

Module 5: Explanations People Can Use

Turning technical explanations into notices, reasons and recourse.
3 lessons
  1. Adverse Action Reason Codes for Credit Decisions
  2. Writing Decision Explanations for GDPR Article 22 Requests and US Adverse-Outcome Notices
  3. Contestability and Recourse: Designing Appeals and Human Review

Module 6: Standards, Documentation and Production Monitoring

Making responsible AI work auditable and keeping it true after launch.
4 lessons
  1. NIST SP 1270, ISO/IEC TR 24027 and ISO/IEC TS 6254 in Practice
  2. Model Cards, Datasheets and Fairness Reports That Hold Up to Review
  3. Monitoring Fairness in Production: Subgroup Drift and Outcome Feedback
  4. Case Review: Assessing a Hypothetical Resume Screening Model

Exam and certificate

The exam unlocks once you have completed every lesson. It draws 20 questions from a larger bank, spread across the modules, with the answer options shuffled for each attempt. You have 45 minutes and need 80% to pass. After a failed attempt you can try again after 24 hours, up to 3 attempts in total. Feedback is given per module, never as the answer key.

Passing earns a certificate of completion and assessment from the Wirewalk Institute of AI with a unique ID that anyone can check at institute.wirewalk.com/verify. It is not an accredited qualification or a vendor certification. How certification works.

The Wirewalk Institute of AI is independent. This course teaches publicly documented products and is not affiliated with or endorsed by Fairlearn, AI Fairness 360, SHAP, or LIME. Product and company names are trademarks of their respective owners.

More in Governance, Risk and Compliance

All 14 courses in this track
gov-001 Free intro Available

AI Rules in Plain English: What Everyone Should Know

Start with existing US law, learn when state and sector requirements need review, and distinguish voluntary frameworks from binding obligations. Practice checking an AI-assisted task and raising a useful concern.

2 h · 7 lessons Free intro
gov-101 Basic Waitlist

AI Compliance Essentials for Every Employee

A plain-language course on using AI at work without creating legal, privacy or reputational problems. Covers the rules that already apply to your job, the difference between approved and shadow tools, what data you may paste, how to check and disclose AI output, and how to recognize and report an AI incident.

5 h · 16 lessons Join the waitlist
gov-201 Intermediate Available

AI Governance and Risk Management: NIST AI RMF and ISO/IEC 42001

A practitioner's course in building AI governance that works: inventory, intake, risk tiering, impact assessment, lifecycle controls and leadership reporting, anchored to NIST AI RMF 1.0, its Generative AI Profile and ISO/IEC 42001. Ends with a 90-day program build plan and ready-to-adapt templates.

14 h · 18 lessons See price at checkout