The use of AI by public authorities.
summary
Panel lists concrete public‑sector AI uses and failures: police facial recognition at Notting Hill Carnival yields 35 false matches and one wrongful arrest; a welfare decision tool is revealed as a spreadsheet of 30 questions, data‑flawed and overturned in court; refugee programs use biometric iris scans in Jordan to deliver services; predictive policing, teacher scoring, autopilot crashes and mail‑sorting examples highlight opaque models, unclear accuracy, hidden feedback loops and weak redress, urging transparency, auditability and strict limits.
AI-written summary · about
chapters (10)
- 0:00Start
- 2:13What Is Ai
- 10:58Drafting Recommendations for the Use of Automated Decision-Making by Public Authorities
- 12:57Public Transit
- 14:25Public Education
- 16:24Main Challenges
- 20:31France
- 28:18The Threshold Question Is It Transparent
- 32:27Five What Kind of Side Effects Does the Algorithm Have
- 37:52Use of Ai Applications To Make More Informed Decisions
related talks
29mResponsible AI Challenges and Recommendations
01h 09mGOVERNANCE AND REGULATION OF AI FROM THE PERSPECTIVE OF AUTONOMY AND PRIVACY
01h 16mBEYOND ETHICS WASHING:IMPACT ASSESSMENTS, AUDITS, AND OVERSIGHT FOR AI
29mThe Political Roots of AI Policy
41mAI and the Law
21mCapabilities and Challenges of AI in Digital Governance
47mEthics & AI: Identifying & preventing bias in predictive models
35mFrom hype to responsibility: what works & matters for whom in data and AI?