Chapters

• When computers say no: Addressing algorithmic bias through EU anti-discrimination law
• Outline
• 1. How do algorithmic machines discriminate?
• The problem
• Examples
• AI’S BROKEN PROMISES: VIGNETTE I: FACIAL RECOGNITION IN EDUCATION
• AI’S BROKEN PROMISES: VIGNETTE II: RISK SCORING IN THE EMPLOYMENT MARKET
• AI’S BROKEN PROMISES: VIGNETTE III: FRAUD DETECTION AND THE WELFARE STATE (1/2)
• AI’S BROKEN PROMISES: VIGNETTE III: FRAUD DETECTION AND THE WELFARE STATE (2/2)
• AI’S BROKEN PROMISES: VIGNETTE IV: PROFILING AND AD TARGETING
• Open AI's Chat GPT & DALLE-2: algorithmic stereotyping
• 'Garbage in, garbage out'
• 2. What is unlawful discrimination under EU law?
• A regulatory puzzle: EU anti-discrimination law
• When is bias unlawful discrimination?
• 3. How to redress algorithmic discrimination?
• Legal consequences of the direct/indirect qualification
• Shift of the burden of proof
• Effective remedies
• What remedies?
• 4. The interaction between EU anti-discrimination law, data protection and AI regulation
• WHAT DOES THE LAW SAY ABOUT ALGORITHMIC BIAS?
• Proposal for a Regulation laying down harmonised rules on artificial intelligence (EU AI Act)
• Zooming on the EU AI Act
• The EU AI Act: requirements for data
• The GDPR and 'sensitive data'
• The EU AI Act and the GDPR
• Article 9
• 5. Shortcomings and remaining issues
• Redressing algorithmic discrimination: detection and enforcement
• EU AIA, AILD and ADL: do they fit together?