{"schema_version":"0.10.0","id":"LR-2026-0354","record_version":1,"title":"UK police facial recognition shown to have racial bias for Black and Asian subjects","verification":{"human_verified":false,"method":"automated_gate","auto_verified":true,"score":51.9,"score_breakdown":{"source_authority":12.1,"independent_corroboration":15.0,"implicated_party_confirmed":0.0,"citation_integrity":14.0,"specificity":0.0,"completeness":3.2,"unhedged_language":5.0},"checks":{"schema_valid":true,"every_field_cited":true,"quotes_verbatim":true,"source_fetched":true,"distinct_incident":true,"dates_plausible":true,"sources_agree":true,"source_describes_incident":true,"severe_backed_by_source":true,"record_integrity":true,"eligible":true,"about_this_case":true},"gate_version":"gate-2026-10-18.7","cross_check_model":null,"cross_check_agreement":null,"duplicate_of":null,"verified_by":null,"verified_at":"2026-11-01","extraction_run_id":null,"evidence_level":{"level":"confirmed","reason":"2 independent outlets","original_document":null,"reports_total":3,"reports_independent":2,"reports_derivative":1,"developing":false,"disputed":false,"rules_version":"evidence-2026-10-29.1","computed_at":"2026-11-01","history":[{"date":"2026-11-01","from_level":null,"to_level":"confirmed","reason":"2 independent outlets"}]},"sources_hash":"3f4d08f870960d5c00324db87f8a9d101c161aa4e879eeae92e97b562dbcb197"},"system":{"model_name":null,"model_version":null,"developer":"unknown","harness":null,"tools_granted":[],"permissions_scope":null,"autonomy_level":"unknown","guardrails_present":[],"deployment_context":"production","tier":null,"lifecycle_stage":"production","integration_type":null,"intended_use":"Catching serious offenders","safeguard_state":null,"production_intended":null},"event":{"date_occurred":null,"date_disclosed":"2025-12-05","incident_class":"bias_disparate_impact","failure_class":null,"agent_involved":null,"secondary_failure_classes":[],"trigger":"unknown","duration":null,"detection_method":"unknown","summary":"The Home Office's retrospective facial recognition tool, used by UK police to match images against the police national database of known or wanted criminals, was tested by the National Physical Laboratory (NPL). The testing found it was more likely to incorrectly match Black and Asian people than white people. The false positive identification rate for white subjects was 0.04%, against 4.0% for Asian subjects and 5.5% for Black subjects. The Home Office admitted the bias after the testing. The UK data protection watchdog asked the Home Office for urgent clarity, and police and crime commissioners urged caution over national expansion. The Home Office said it had procured a new algorithm with no statistically significant bias and asked the police inspectorate and forensic science regulator to review use of the technology.","tags":[],"date_detected":null,"near_miss":null,"root_cause":null,"root_cause_detail":null,"failure_class_detail":null,"responsible_party":null,"adversarial_actor":null,"input_provenance":null,"time_to_contain":null,"operations_paused":null,"remediation":null,"precursor_events":null,"primary_risk":"RK16"},"impact":{"direct_loss_usd":null,"remediation_hours":null,"records_exposed":null,"third_parties_affected_count":null,"severity":"major","severity_because":"Test: regulator involved. Fact: the UK data protection watchdog (ICO) asked the Home Office for urgent clarity over racial bias before considering next steps.","harm_class":[],"affected_description":null,"protected_class_involved":null,"data_categories_exposed":[],"loss_range_usd":null,"reversibility":null,"downstream_exposure_count":null},"evidence":{"sources":[{"url":"https://futurism.com/artificial-intelligence/police-uk-surveillance-ai","publisher":"Futurism","date":"2025-12-05","type":"press","title":"Police Admit AI Surveillance Panopticon Still Has Issues With \"Some Demographic Groups\" - Futurism","archived_url":null,"attachment":"first_report","independence":"independent"},{"url":"https://www.theguardian.com/technology/2025/dec/05/urgent-clarity-sought-over-racial-bias-in-uk-police-facial-recognition-technology","publisher":"The Guardian","date":"2025-12-05","type":"press","title":"‘Urgent clarity’ sought over racial bias in UK police facial recognition technology - The Guardian","archived_url":null,"independence":"independent"},{"url":"https://www.theguardian.com/technology/2025/dec/05/home-office-facial-recognition-tech-issue-black-asian-subjects","publisher":"The Guardian","date":"2025-12-06","type":"press","title":"Home Office admits facial recognition tech issue with black and Asian subjects - The Guardian","archived_url":null,"independence":"same_owner"}],"per_field_confidence":{"system.intended_use":"C","system.deployment_context":"C","system.lifecycle_stage":"C","event.incident_class":"C","impact.severity":"D","impact.severity_because":"D","event.primary_risk":"D","event.summary":"C","system.autonomy_level":"D","event.trigger":"D","event.detection_method":"D","context.country":"D","system.developer":"D","event.date_disclosed":"D","context.denominator_available":"D"},"field_citations":{"system.intended_use":[{"source_index":1,"quote":"The report revealed that the technology, which is intended to be used to catch serious offenders, is more likely to incorrectly match black and Asian people than their white counterparts."}],"system.deployment_context":[{"source_index":0,"quote":"“The language is technical, but behind the detail it seems clear that technology has been deployed into operational policing without adequate safeguards in place.”"}],"system.lifecycle_stage":[{"source_index":0,"quote":"“The language is technical, but behind the detail it seems clear that technology has been deployed into operational policing without adequate safeguards in place.”"}],"event.incident_class":[{"source_index":1,"quote":"Analysts who examined the police national database’s retrospective facial recognition technology tool at a lower setting found that “the false positive identification rate (FPIR) for white subjects (0.04%) is lower than that for Asian subjects (4.0%) and black subjects (5.5%)”."}],"impact.severity":[{"source_index":1,"quote":"The UK’s data protection watchdog has asked the Home Office for “urgent clarity” over racial bias in police facial recognition technology before considering its next steps."}],"impact.severity_because":[{"source_index":1,"quote":"The UK’s data protection watchdog has asked the Home Office for “urgent clarity” over racial bias in police facial recognition technology before considering its next steps."}],"event.summary":[{"source_index":1,"quote":"Analysts who examined the police national database’s retrospective facial recognition technology tool at a lower setting found that “the false positive identification rate (FPIR) for white subjects (0.04%) is lower than that for Asian subjects (4.0%) and black subjects (5.5%)”."},{"source_index":1,"quote":"The Home Office has admitted that the technology was “more likely to incorrectly include some demographic groups in its search results”, after testing by the National Physical Laboratory (NPL) of its application within the police national database."},{"source_index":1,"quote":"The UK’s data protection watchdog has asked the Home Office for “urgent clarity” over racial bias in police facial recognition technology before considering its next steps."},{"source_index":2,"quote":"A new algorithm has been independently tested and procured, which has no statistically significant bias."},{"source_index":1,"quote":"Given the importance of this issue, we have also asked the police inspectorate, alongside the forensic science regulator, to review law enforcement’s use of facial recognition."}]},"notes_on_gaps":"The sources do not state the system developer, the event date, the detection date, the disclosure date, the number of people affected or wrongly matched, records exposed, third parties affected, or any financial loss, so those fields are left out. The system is a facial recognition tool, not an autonomous agent, so agent_involved is false and no failure_class is given. No trigger or root cause is stated, so those are left out. Responsible party is not stated, so it is left out. The bias test was run by the National Physical Laboratory; I treated the Home Office admission as the detection method 'third_party'. The B1 control outcome 'held' rests on the Home Office's own claim that manual visual checks exist; no source independently confirms it operated. Severity 'major' is based on the regulator (ICO) being involved, not on a stated count of affected parties. Source 0 gives the NPL figures as 0.04 percent, 4.0 percent and 5.5 percent, while sources 1 and 2 give the same figures in the same form. The Home Office said it had not stated a date for the new algorithm test beyond 'early next year'. [case created by the backfill: read by claude-haiku-5-5, backfill-2026-11-01.2, 2026-11-01; not yet reviewed] [2026-11-01: evidence level set: Confirmed (2 independent outlets)]"},"context":{"industry":null,"org_size_band":null,"country":"unknown","denominator_available":false,"denominator":null,"related_incident_ids":[],"systemic_exposure":null,"shared_model_id":null,"upstream_incident_id":null,"incident_evidence":null,"regulatory":null,"coverage":null}}