Every record sourced from law enforcement booking, traceable to a booking-based arrest record
Every identity in UMbRA has a documented origin. No scraped images. No aggregated public records
UMbRA records meet the evidentiary standard required for prosecution and adjudication
UMbRA: Data Provenance and Chain of Custody
The accuracy of any identification system depends on the quality and provenance of its reference data. UMbRA contains 55.5 million identities sourced directly from law enforcement booking records, each linked to a booking-based arrest. Every record has a documented chain of custody. This is the difference between an identification and a guess.
Why Data Provenance Determines System Value
A facial identification system is only as reliable as its reference database. If the reference data is unverified, the identifications are unverified. If the reference data has no chain of custody, the identifications have no evidentiary value. UMbRA is the only identity database built entirely on booking-verified law enforcement booking records with complete provenance documentation.
-
Arrest and Booking
An individual is arrested and booked by a law enforcement agency. During booking, identity records are captured and a booking photograph is taken. The identity records establish a verified identity.
-
Identity Verification
The booking identity records are verified against law enforcement records (law enforcement identity verification systems) and verified against existing records. This establishes a one-to-one link between the physical person and their identity record.
-
Record Ingestion
The booking record (photograph, identity data, charge information) is ingested into UMbRA. The identity verification link is preserved. Each record is traceable to its source agency, booking date, and identity match.
-
Operational Use
When RTIS performs a comparison, it compares the captured image against UMbRA records. A match links the detected person to a booking-verified identity with a documented law enforcement source. This chain of custody is what makes the identification court-ready.
Data Source Comparison
| Feature | UMbRA | Scraped FR Databases | Data Aggregators |
|---|---|---|---|
| Source | Law enforcement booking records | Social media platforms (scraped without consent) | Public records, commercial databases, credit bureaus |
| Verification method | Booking-based identity confirmation | None. Image-to-name association from social media profile. | Name/address matching. No biometric verification. |
| Chain of custody | Complete. Every record traceable to source agency and booking event. | None. Images scraped at scale with no provenance documentation. | Partial. Aggregated from multiple sources with varying quality. |
| Record count | 55.5M verified identities | 50B+ images (Clearview, self-reported) | Varies by vendor. Hundreds of millions of records. |
| Court readiness | Meets evidentiary standard for prosecution. | Clearview's own contracts disclaim admissibility. An Ohio judge dismissed a murder case based on Clearview results. | Aggregated data regularly challenged in litigation. Data staleness and error rates undermine reliability. |
| Consent and legality | Law enforcement booking records collected under lawful authority. | Scraped without user consent. Banned by Meta, Google, Twitter. Fined by UK, France, Italy, Australia. | Varies. Consumer data use subject to FCRA, DPPA, and state privacy laws. |
Database size is meaningless without data provenance. 50 billion unverified images are worth less than 55.5 million verified identities when the identification must hold up in court.
Request a Data Provenance Briefing
Schedule a session with our data operations team to review UMbRA's sourcing methodology, chain of custody documentation, and evidentiary standards. Available under NDA.