CAID facial matching helps Scottish child abuse investigators

Scotland's Biometrics Commissioner and police inspectorate found that CAID facial matching made some contribution to child abuse investigations and staff wellbeing.

  • 1 an organization

    Staff told the reviewers that the review of a 10,000-image case fell from up to three days to about an hour.

  • Not rated a person

    The reviewers found a contribution to the wellbeing of staff who view abuse imagery but give no measure of it.

  • Not rated society

    The review reports no count of children identified or safeguarded through Police Scotland's use of the tool.

Depth runs from 1, a convenience or time saved, to 5, a life or a livelihood preserved.

Status
Reported benefit.
Evidence

foundstudiedreportedclaimed

An independent audit report supports the benefit.

Compared with

Manual review of a seized device's images before facial matching. Assumed without a measurement.

The system’s part
Not isolated. The reviewers relied on interviews, case studies and search counts because Police Scotland collects no data on the tool's effectiveness. They make no comparison of investigations with and without facial matching. The review attributes the matching to NEC's face recognition algorithms and does not describe how they were trained.
Limits
Not yet checked.
Costs and harms
Not yet checked.
Later
Not yet checked.

Download data Cite Suggest a correction

Benefit type
Reduced Task Time
System
Facial matching function of the UK Child Abuse Image Database (CAID)
Made by
NEC (face recognition technology, per the review), within the Home Office's Child Abuse Image Database, which West Yorkshire Police manages
Used by
Police Scotland (Cybercrime Investigations, Digital Forensics and the Victim ID Unit)
Where
Scotland, United Kingdom
Stage
In routine use
Date
, when the source was published
Kind
Aggregate result
Broad benefit
Identifying Objects, Actions, and Events, Processing Information
Collections
Scrutiny and findings, Cross-border records
Last revised
31 October 2026

Sources

  1. Audit report, by authors with no tie to the system.

    Scottish Biometrics Commissioner and HM Inspectorate of Constabulary in Scotland. Joint Assurance Review of the use of retrospective facial search technologies for criminal justice and police purposes in Scotland (SBC/2025/02). Laid before the Scottish Parliament in March 2025. Summary of findings, section on effectiveness.

    CAID facial matching has made some contribution to improving the effectiveness of investigations into online child sexual abuse and exploitation crimes and ensuring staff and officers' wellbeing working in this area.
    For instance, we were told that a case involving 10,000 images that previously required up to three days to review can now be processed in about an hour using CAID, before final human supervision.

Details

AI system

Workflow

CAID runs facial matching automatically on every new case and across historic cases. Vetted and trained operators review potential matches, and every decision involves human verification. Cybercrime Investigations and Digital Forensics load images from devices seized under warrant, and the Victim ID Unit investigates tagged victims.

Follow-up

Conditions for replication

The tool is used only on seized devices that the cybercrime unit refers when a child is involved. Human verification stays part of every decision.

Entry details

Sector

Policing.

Purpose

NEC built the face recognition technology that the Child Abuse Image Database uses to match faces across seized images and help police identify victims and offenders.

Not yet checked: model, release, provider, role, configuration, inputs and outputs, autonomy, human oversight, AI contribution, beneficiaries, reach, distribution and access, duration, every measurement question, every money question, ownership and control, dependencies, cross-border connections, geopolitical implications, replication attempts, availability, institutional decision, grounds, action and result, success criterion, every evidence question, observed or intended, alternative assessments, related entries, other identifiers, category mappings.

Cite this record

Peter H. Vartanian. (2026). CAID facial matching helps Scottish child abuse investigators (AIBD-0883). Artificial Intelligence Benefits Database, edition of 31 October 2026. https://petervartanian.xyz/aibd/records/AIBD-0883.html (archived at https://doi.org/10.7910/DVN/AADSII, version 4.0)

BibTeX Data Field definitions