Is My AI Racist? Or How Not To Break The Law With Machine Learning
summary
Covers how GDPR already restricts automated decisions, arguing Article 22 GDPR bars fully automated individual decisions unless contract necessity, national law, or express consent applies; recommends using synthetic data when anonymization breaks utility; cites telematics insurance that adjusts premiums in real time and a bank error that set withdrawal limits to zero as concrete failure modes; gives a 49/48/4 percent heuristic for noncompliance, circumvention, and fully compliant cases; insists on right to human intervention and explainability vs trade secrets.
AI-written summary · about
chapters (7)
- 0:00Introduction
- 1:33Generating Business Value
- 2:24Data Collection
- 12:45Discrimination
- 18:15GDPR
- 25:10Fully automated decisions
- 31:34Human intervention
related talks
30mAI Regulation - How to implement AI legally compliant?
29mResponsible AI Challenges and Recommendations
26m(Un)ethical Artificial Intelligence: How to Keep the AI Fair for Everyone
52mSome pitfalls in AI
01h 09mGOVERNANCE AND REGULATION OF AI FROM THE PERSPECTIVE OF AUTONOMY AND PRIVACY
01h 19mAligning Data Privacy and AI Governance
01h 17mBias, Discrimination and Data Protection in the Age of GPAI
54mEthical Machine Learning