Document and Evaluate AI Ethics Course

Lesson 1 • Core Ethical Theories and AI

Surveys consequentialism, deontology, virtue ethics, and contractualism as applied to AI decisions. Provides the theoretical lenses used throughout the course.

Lesson 2 • Historical Context of AI Ethics

Traces ethical debates from early computing through modern machine learning. Shows how past controversies shaped current principles and regulatory expectations.

Lesson 3 • What Is AI Ethics

Defines AI ethics as a discipline and distinguishes it from AI safety and AI law. Anchors the chapter by establishing shared terminology for all subsequent topics.

Lesson 1 • Defining Fairness Mathematically

Introduces demographic parity, equalized odds, calibration, and individual fairness metrics. Demonstrates why no single metric satisfies all fairness goals simultaneously.

Lesson 2 • Sources of Bias in AI

Categorizes bias as historical, representation, measurement, and aggregation types. Connects each source to concrete failure modes students will analyze later.

Lesson 3 • Bias Mitigation Strategies

Presents pre-processing, in-processing, and post-processing debiasing techniques. Students evaluate trade-offs between accuracy and fairness for each approach.

Lesson 4 • Fairness in High-Stakes Domains

Applies fairness concepts to hiring, lending, healthcare, and criminal justice contexts. Reinforces that domain context shapes which fairness criteria are most appropriate.

Lesson 1 • Communicating Explanations to Stakeholders

Teaches translation of technical explanations into plain-language narratives for non-expert audiences. Connects to stakeholder mapping from Chapter 1.

Lesson 2 • Transparency vs. Explainability

Distinguishes system transparency, model interpretability, and post-hoc explainability as separate concepts. Clarifies which concept addresses which accountability need.

Lesson 3 • Limits and Risks of Explainability

Addresses explanation manipulation, false confidence, and the privacy risks of revealing model internals. Prepares students to use explainability responsibly.

Lesson 4 • Intrinsically Interpretable Models

Examines decision trees, linear models, and rule-based systems as inherently transparent architectures. Students weigh interpretability against predictive performance trade-offs.

Lesson 1 • Privacy-by-Design in AI Pipelines

Applies privacy-by-design principles across data collection, model training, and deployment stages. Students produce a privacy impact assessment for a sample pipeline.

Lesson 2 • Re-identification and Inference Risks

Demonstrates how anonymized data can be re-identified and how models leak sensitive attributes. Reinforces why technical anonymization alone is insufficient.

Lesson 3 • Privacy-Enhancing Technologies

Covers differential privacy, federated learning, and synthetic data as technical privacy safeguards. Students evaluate each technology's privacy-utility trade-off.

Lesson 4 • Privacy Principles in AI Contexts

Introduces data minimization, purpose limitation, storage limits, and contextual integrity. Grounds privacy analysis in principles applicable across regulatory regimes.

Lesson 1 • AI Policies and Internal Standards

Guides students in drafting AI use policies, acceptable-use standards, and escalation procedures. Connects internal standards to external regulatory expectations.

Lesson 2 • Accountability Concepts in AI

Defines answerability, liability, and redress as distinct accountability dimensions. Establishes the conceptual vocabulary for designing governance structures.

Lesson 3 • Incident Response and Remediation

Establishes protocols for detecting, reporting, and remediating AI-related harms. Students draft an incident response plan using a structured template.

Lesson 4 • Organizational AI Governance Models

Compares centralized ethics boards, distributed ownership, and hybrid governance models. Students assess which model fits different organizational sizes and risk profiles.

Lesson 1 • Risk Communication and Reporting

Teaches how to communicate AI risk findings to executives, boards, and regulators clearly. Reinforces accountability structures from Chapter 5 with practical reporting formats.

Lesson 2 • Ongoing Monitoring and Model Drift

Addresses performance degradation, distribution shift, and concept drift as post-deployment risks. Students design a monitoring dashboard with alert thresholds.

Lesson 3 • Pre-Deployment Risk Assessment

Covers threat modeling, red-teaming, and pre-launch checklists for AI systems. Students conduct a structured pre-deployment assessment on a sample system.

Lesson 4 • Risk Tiering and Prioritization

Introduces risk tiering frameworks that classify AI systems by potential harm severity and breadth. Students apply tiering to determine appropriate oversight intensity.

Lesson 1 • Designing an Ethics Evaluation Plan

Guides students through scoping, criteria selection, evidence gathering, and scoring design. Produces a reusable evaluation plan template applicable across AI project types.

Lesson 2 • Quantitative Ethics Metrics

Operationalizes fairness, privacy, and robustness as measurable metrics with defined thresholds. Connects to bias metrics from Chapter 2 and privacy tools from Chapter 4.

Lesson 3 • Documenting and Reporting Evaluations

Establishes standards for writing evaluation reports, model cards, and datasheets for datasets. Students produce a complete evaluation report for a sample AI system.

Lesson 4 • Qualitative Ethics Assessment Techniques

Covers expert panels, ethics red teams, stakeholder interviews, and scenario analysis. Balances quantitative metrics with contextual human judgment.

Lesson 1 • Measuring Ethics Program Maturity

Introduces maturity models for AI ethics programs and key performance indicators for ethics outcomes. Students assess an organization's current maturity and set improvement targets.

Lesson 2 • Building an Ethical AI Culture

Addresses leadership behaviors, psychological safety, and incentive structures that sustain ethics culture. Students diagnose cultural barriers and design targeted interventions.

Lesson 3 • Ethics Training and Capability Building

Designs role-specific ethics training programs for engineers, product managers, and executives. Students create a capability-building roadmap for a sample organization.

Lesson 4 • Embedding Ethics in Product Lifecycles

Integrates ethics checkpoints into agile sprints, design reviews, and launch gates. Connects evaluation methods from Chapter 7 to product development workflows.