Overview
Biometrica Systems, Inc. integrates limited forms of Artificial Intelligence (AI) exclusively to support public safety, victim recovery, and investigatory efforts. Our use of AI is designed around privacy, human oversight, and compliance with applicable international, federal, and state laws and regulatory frameworks.
We do not deploy or utilize:
- Generative AI
- Predictive policing models
- Mass surveillance technologies
- Automated decision-making without human review
Biometrica’s AI use is strictly agentic and subject to strong, documented privacy and ethics safeguards.
Purpose of AI in Biometrica’s Systems
Biometrica incorporates AI into the following privacy-protective, public-safety-specific systems:
| System | Purpose | AI Role | Compliance Notes |
|---|---|---|---|
| RTIS (Real-Time Threat Identification System) | Detects individuals with active, law-enforcement-verified threat histories | Facial similarity scoring performed by an isolated third-party NIST-evaluated provider | Subject to human verification before alert issuance |
| RVIS (Real-Time Victim Identification System) | Identifies missing or endangered persons based on law enforcement or quasi-law enforcement records | Facial similarity scoring via third-party NIST-evaluated provider | Subject to human verification before alert issuance |
| QAPLA | Provides face similarity verification for investigative support (e.g., eliminating false leads) | Non-retentive similarity scoring tool | Used only by credentialed law enforcement or investigative personnel |
| eMotive | Continuous FCRA-compliant employee background check monitoring for new criminal events | Criminal event notifications (no facial comparison) | Employee consent required; FCRA and U.S. state law compliant |
How Biometrica Uses AI Responsibly
Human-in-the-Loop
Biometrica mandates human oversight at every stage where AI is used. No AI-generated output, including similarity scores, triggers automated actions without review by a trained human analyst.
Purpose-Limited
Biometrica’s AI is only used to:
- Support law enforcement and public safety missions
- Protect vulnerable individuals (e.g., missing persons, trafficked individuals)
- Enable investigatory accuracy and speed without compromising privacy
We do not engage in:
- Behavioral profiling
- Predictive analytics for policing
- Social scoring or categorization
- General public surveillance
Data Minimization
Biometrica’s AI usage is strictly aligned with data minimization principles:
- We process only images required for immediate comparison (no video)
- No biometric identifiers, faceprints, or templates are retained or accessible to Biometrica staff
- Matched alerts generate anonymized, human-verified records without biometric metadata
Technical & Organizational Safeguards
- NIST-Approved Independent Provider: Facial similarity scoring is performed by a third-party, NIST-evaluated and approved provider in a secure, isolated, black box environment.
- No Biometric Library: Biometrica does not maintain, access, or contribute to any biometric gallery.
- Immutable Audit Trails: All relevant activity is subject to immutable chain-of-custody logging for law enforcement integrity.
- FCRA Compliance: eMotive adheres to the U.S. Fair Credit Reporting Act for continuous criminal monitoring.
Regulatory Compliance Alignment
Biometrica’s AI design and operation meet the requirements of:
- GDPR (General Data Protection Regulation)
- UK Data Protection Act (DPA) 2018
- EU AI Act (2024)
- Canadian PIPEDA & Quebec Law 25
- All U.S. Federal & State Privacy Laws
- FCRA (U.S. Fair Credit Reporting Act)
- Applicable Sector-Specific Laws (FERPA, HIPAA, etc.)
Summary
Biometrica’s AI use is minimal and the systems in use have a specific, limited purpose:
- Protect public safety without engaging in mass surveillance
- Support investigations while respecting individual rights
- Operate under full human control and oversight
- Are fully auditable and privacy-first by design
For further inquiries, please contact:
privacy@biometrica.com
Last Updated: March 2025
