Brazil — 1 Million Schoolchildren Scanned Daily by Facial Recognition Across 1,700+ Schools (2026)

Incident Summary

Brazil’s Paraná state deployed facial recognition technology from Slovak company Innovatrics across more than 1,700 schools, scanning approximately 1 million children daily to track attendance.^[1] The Innovatrics technology had been rejected by the European Union, but was exported to Brazil where regulatory restrictions on biometric surveillance of children are less developed.^[2] The system achieved only 91.1% accuracy — below the 95% threshold generally considered minimum for biometric systems — meaning that nearly 9 out of every 100 scans produce incorrect results, a rate that at the scale of 1 million daily scans translates to approximately 89,000 errors per day.^[1] The facial recognition data feeds into welfare eligibility determinations, creating a direct link between biometric surveillance of children and families’ access to social benefits. A Brazilian prosecutor has challenged the system under the country’s data protection law, arguing that the mass biometric surveillance of minors violates their privacy rights.^[3]

Key Facts

• Scale: 1 million children scanned daily across 1,700+ schools^[1]
• Technology: Innovatrics (Slovak company, rejected by EU)^[2]
• Accuracy: 91.1% — below 95% threshold^[1]
• Welfare link: FRT attendance data feeds welfare eligibility^[1]
• Legal challenge: Prosecutor challenged under data protection law^[3]
• Location: Paraná state, Brazil

Threat Patterns Involved

Primary: Mass Surveillance Amplification — The deployment of facial recognition across 1,700+ schools to scan 1 million children daily represents mass biometric surveillance of minors at a scale that would not be feasible without automated systems, amplifying the state’s ability to monitor children’s movements and attendance.

Secondary: Biometric Exploitation — The collection of children’s facial biometric data — from a population unable to consent — for use in welfare eligibility determinations exploits biometric data for purposes beyond its stated attendance-tracking function.

Significance

1. EU-rejected tech exported to Global South — The deployment of technology rejected by the EU in Brazilian schools demonstrates how surveillance technology banned in one jurisdiction can be exported to countries with less restrictive regulatory frameworks, creating a two-tier global system of privacy protection
2. Children as surveillance subjects — The daily facial scanning of 1 million schoolchildren normalizes biometric surveillance from childhood, creating a generation habituated to state biometric monitoring
3. 91.1% accuracy at scale — At 1 million daily scans, the 8.9% error rate translates to approximately 89,000 daily errors, meaning tens of thousands of children are incorrectly identified or missed every day with potential consequences for their families’ welfare benefits
4. Welfare-surveillance link — Connecting facial recognition attendance data to welfare eligibility creates a coercive dynamic where families must submit their children to biometric surveillance to maintain access to social benefits