Incidente 829: Sistema de reconocimiento facial en Buenos Aires activa controles policiales basados en coincidencias falsas
Descripción: El sistema de reconocimiento facial de Buenos Aires marcó erróneamente a personas inocentes como delincuentes, lo que provocó detenciones indebidas. Investigaciones judiciales indican que la tecnología podría haber sido utilizada indebidamente para la vigilancia y la recopilación de datos no autorizadas. A pesar de los riesgos para la privacidad, el sistema se ha utilizado ampliamente sin divulgar completamente sus estándares ni salvaguardas.
Editor Notes: Reconstruction of the timeline of events: (1) 2019: Buenos Aires implements a facial recognition system aimed at enhancing public safety, capturing thousands of individuals. (2) After implementation in 2019: At least 140 individuals, including Guillermo Ibarrola, are erroneously flagged as criminals due to database errors, leading to police checks and detentions. (3) 2020: The facial recognition feature is deactivated as a precaution during the COVID-19 pandemic and remains off by judicial order. (4) December 2023: Journalists confirm that their biometric data was accessed, which in turn prompted further scrutiny by them. (5) February 5, 2024: The Pulitzer Center publishes a report on the issues surrounding Buenos Aires's facial recognition system.
Entidades
Ver todas las entidadesAlleged: Government of Argentina developed an AI system deployed by Government of Argentina , Government of Buenos Aires y Argentinean Ministry of Security, which harmed Argentinean citizens , Buenos Aires residents y Guillermo Ibarrola.
Estadísticas de incidentes
Risk Subdomain
A further 23 subdomains create an accessible and understandable classification of hazards and harms associated with AI
2.1. Compromise of privacy by obtaining, leaking or correctly inferring sensitive information
Risk Domain
The Domain Taxonomy of AI Risks classifies risks into seven AI risk domains: (1) Discrimination & toxicity, (2) Privacy & security, (3) Misinformation, (4) Malicious actors & misuse, (5) Human-computer interaction, (6) Socioeconomic & environmental harms, and (7) AI system safety, failures & limitations.
- Privacy & Security
Entity
Which, if any, entity is presented as the main cause of the risk
AI
Timing
The stage in the AI lifecycle at which the risk is presented as occurring
Post-deployment
Intent
Whether the risk is presented as occurring as an expected or unexpected outcome from pursuing a goal
Unintentional
Informes del Incidente
Cronología de Informes
El setenta y cinco por ciento de la zona de la capital argentina está bajo videovigilancia, algo que el gobierno anuncia orgullosamente en vallas publicitarias. Pero el sistema de reconocimiento facial, parte de la extensa infraestructura d…
Variantes
Una "Variante" es un incidente de IA similar a un caso conocido—tiene los mismos causantes, daños y sistema de IA. En lugar de enumerarlo por separado, lo agrupamos bajo el primer incidente informado. A diferencia de otros incidentes, las variantes no necesitan haber sido informadas fuera de la AIID. Obtenga más información del trabajo de investigación.
¿Has visto algo similar?
Incidentes Similares
Did our AI mess up? Flag the unrelated incidents
Northpointe Risk Models
· 15 informes
Predictive Policing Biases of PredPol
· 17 informes
Uber AV Killed Pedestrian in Arizona
· 25 informes
Incidentes Similares
Did our AI mess up? Flag the unrelated incidents
Northpointe Risk Models
· 15 informes
Predictive Policing Biases of PredPol
· 17 informes
Uber AV Killed Pedestrian in Arizona
· 25 informes