AI Fairness and Social Impact

Edit

• Chapter 1.1: What is Fairness in AI? Definitions and Context
• Chapter 1.2: Discrimination, Equity, and Power in Algorithmic Systems
• Chapter 1.3: Key Concepts: Group Fairness, Individual Fairness, Procedural Fairness
• Chapter 1.4: Historical Case Studies of Bias and Harm in AI Deployment

• Chapter 2.1: Popular Fairness Metrics and When to Use Them
• Chapter 2.2: Trade-offs Between Accuracy, Fairness, and Utility
• Chapter 2.3: Technical vs. Contextual Fairness
• Chapter 2.4: Critical Perspectives: Fairwashing, Audit Theatre, and Incomplete Solutions

• Chapter 3.1: Criminal Justice Algorithms and Racial Disparities
• Chapter 3.2: AI in Health, Education, and Public Benefits
• Chapter 3.3: Labor Market Impacts and Discrimination in Hiring Tools
• Chapter 3.4: Surveillance, Policing, and AI at the Margins

• Chapter 4.1: Who Gets to Define Fairness? Power and Representation
• Chapter 4.2: Case Studies of Participatory AI Design
• Chapter 4.3: Impact Assessments and Community Consultations
• Chapter 4.4: Building Culturally Responsive AI Systems

• Chapter 5.1: Policy Responses and Legislative Proposals (EU AI Act, Algorithmic Accountability Act, etc.)
• Chapter 5.2: Organizational Governance: AI Ethics Boards and Equity Audits
• Chapter 5.3: Transparency, Documentation, and Accountability Mechanisms
• Chapter 5.4: Interventions for Equitable AI: Redesign, Rejection, Regulation

• Chapter 6.1: Evaluating Real-World AI Systems for Fairness and Harm
• Chapter 6.2: Capstone Project: Social Impact Assessment of an AI System
• Chapter 6.3: Presenting Findings to Stakeholders (Public, Technical, and Policy)
• Chapter 6.4: Leadership for AI Justice: Advocacy, Allyship, and Institutional Change

Video content aligned with weekly modules

Foundations of Fairness in AI

• Introduction to AI Fairness: Historical and Technical Context
• Definitions and Dimensions: Group vs. Individual vs. Procedural Fairness
• Recognizing Discrimination in Algorithmic Decision-Making
• Technical Fairness Metrics: Equalized Odds, Demographic Parity, Calibration
• Fairness-Utility Trade-offs: What Should Be Optimized?
• Critical Views on Fairwashing and Ethics Theater
• Case Study: COMPAS and Racial Bias in Criminal Justice Algorithms
• Week 1 Summary: Why Definitions of Fairness Matter

Social Impacts of AI Across Sectors

• AI in Criminal Justice: Risk Assessments and Sentencing Tools
• Algorithmic Bias in Healthcare and Diagnosis Tools
• Discrimination in Automated Hiring Systems
• Education Algorithms: Admissions, Placement, and Funding
• Surveillance AI: Facial Recognition and Marginalized Communities
• The Role of Affected Communities in Fair AI Design
• Participatory AI: Community-Led Technology Projects
• Week 2 Case Study: Auditing a Public-Sector Algorithm

Governance, Ethics, and Social Change

• Policy Frameworks for Fair AI (EU AI Act, NYC Local Law 144, etc.)
• Building Ethical Review Boards and Equity Audit Processes
• Transparency Tools: Datasheets, Model Cards, and Impact Statements
• AI Documentation and Disclosures: Legal and Moral Duties
• Redesigning or Rejecting AI Systems: Intervention Strategies
• Capstone Guide: Conducting a Social Impact Evaluation
• Communicating Fairness Audits to Stakeholders
• Final Wrap-Up: Sustaining Equity-Centered AI Practices

Live, interactive sessions aligned with weekly themes

Title: Defining Fairness in Practice: Beyond Math, Toward Justice
Duration: 60 minutes
Focus: Exploring the limitations of purely technical definitions of fairness and introducing critical frameworks for social context
Guest: Algorithmic Justice Researcher / Fairness Metrics Scholar
Interactive: Group debate: “Can AI ever be truly fair?” using real-world case examples

Title: AI in the Real World: Harm, Impact, and Community Response
Duration: 75 minutes
Focus: Investigating AI failures in health, policing, labor, and housing—and what accountability looks like
Guest: Civil Rights Advocate / Investigative Journalist on AI
Interactive: Case workshop: Review and discuss a real audit of a public-sector algorithm

Title: Building Accountability: Policy, Participation, and Resistance
Duration: 90 minutes
Focus: Designing systems that prioritize equity, participatory governance, and responsible AI oversight
Guest Panel: Policy Maker + Technologist + Community Organizer
Interactive: Live critique of learner capstone ideas + open forum for stakeholder mapping and communication strategies

Edit

Edit