{"resourceId":"merseyside-police-dpia-300000-custody-images-overdue-deletion-oifr-not-adopted-20261031","versions":[{"version":"external-752d9c558e1d60c3b10444a7a293f39cf83808fc395b491e1667105afb178039","resource":{"id":"merseyside-police-dpia-300000-custody-images-overdue-deletion-oifr-not-adopted-20261031","title":"Merseyside Police impact assessment finds at least 300,000 people with custody images overdue for deletion, and the force did not adopt officer-initiated facial recognition","organization":"Merseyside Police (England)","sector":"Policing / facial recognition and biometric data retention","geography":"Merseyside, England, United Kingdom","publishedAt":"Liverpool Echo, Oct. 31, 2026","publicationDate":"2026-10-31","eventDate":null,"sourceName":"Liverpool Echo","sourceLabel":"Merseyside Police unlawfully holds 300,000 mugshots","sourceUrl":"https://www.liverpoolecho.co.uk/news/liverpool-news/merseyside-police-unlawfully-holds-300000-34834901","evidenceClass":"independent-reporting","outcomeClass":"cautionary","topics":["data-security","governance-procurement","operating-model"],"finding":"The Liverpool Echo reported on Oct. 31, 2026 that an internal Merseyside Police data protection impact assessment (DPIA) on officer-initiated facial recognition (OIFR), a phone tool that checks a photo taken during a stop against police databases, found \"potentially 300,000 nominals who could have multiple images overdue for deletion\" in the Niche records system. The assessment warned the force could compare against an unlawful image. The Echo understands OIFR was never adopted because of this. The force says OIFR is not in operational use and that custody image retention is a long-standing national issue. Big Brother Watch supplied the document. A 2012 High Court ruling found long retention of images of people not convicted unlawful; Police Scotland reportedly reviews and deletes such images.","sledRelevance":"A rare documented case of a privacy impact assessment stopping a police AI tool because the data it would search was not lawfully held.","evidence":"Opened and read the full Liverpool Echo article, which quotes the DPIA directly. The DPIA itself was not published or opened; the 300,000 figure is the force’s own estimate as quoted.","architectureImplications":"A face-matching tool is only as lawful as the gallery behind it; the gallery (Niche custody images) needs retention flags and deletion before any matching service is connected.","governanceImplications":"The DPIA functioned as a gate: it identified unlawful source data and the tool was not adopted. The underlying retention backlog remains unresolved.","securityPrivacyImplications":"Matching against images that should have been deleted creates legal exposure and privacy harm for people never convicted.","caveats":"Single news report based on a leaked internal document supplied by an advocacy group; the DPIA date is not given; England and Wales law differs from U.S. state law, so the lesson transfers by analogy.","streamIds":["public-safety"],"roles":{"sales":"Sales framing: Merseyside shows a police force stopping a facial recognition tool because its own image database held people who should have been deleted. The buyer problem for any agency is that biometric search exposes old records failures. A bounded engagement could audit gallery lawfulness and retention before a tool is procured. Do not claim the tool would have worked or failed; it was never used, and the 300,000 figure is the force’s estimate.","engineering":"Engineering framing: the integration point is the custody image store, not the matching model. Pre-sales should ask whether each image carries outcome and retention metadata, and whether a match service can exclude images past their deletion date. A proof of value would sample records, measure how many images are overdue, and show a query filter that blocks them. Constraints: legacy records systems, national databases outside force control, and incomplete case outcome data.","delivery":"Delivery framing: the work is a retention clean-up and a gate that keeps biometric tools off until it is done. Ownership sits with the records manager and data protection officer, with the chief officer signing off. Acceptance criteria are proposals: a measured backlog, a deletion schedule, an automated retention filter, and a repeat impact assessment before any pilot. Main risks: a backlog too large to clear and pressure to launch the tool anyway."},"roleDetails":{"sales":{"customerProblem":"Biometric tools would search image galleries that may contain records the agency should no longer hold.","stakeholders":["Chief officer or police chief","Data protection or privacy officer","Records management lead","Legal counsel"],"discoveryQuestions":["How many custody images are past their lawful retention date today?","Can the records system link each image to a case outcome?","Has any biometric tool been run against this gallery already?"],"valueHypothesis":"A gallery lawfulness audit lets the agency decide on biometric tools with known legal exposure.","potentialEngagement":"Retention and gallery-lawfulness assessment ahead of any facial recognition procurement.","unsupportedClaims":["That OIFR would have been accurate or useful","That other forces hold similar backlogs at the same scale","That the DPIA has been formally published"]},"engineering":{"fit":"Relevant to any agency considering face matching against booking or custody photos.","architectureIntegration":"Custody image store in the records management system feeding a matching service; retention metadata and query filters are the control points.","prerequisites":["Image-to-outcome linkage","Retention rules per jurisdiction","Deletion audit log"],"constraints":["Legacy records systems","National databases outside local control","Incomplete case outcome data"],"security":"Block matches against images past retention and log every query with officer and reason.","proofOfValue":["Sample audit of overdue images","Working retention filter on a test gallery","Deletion log reconciled to the sample"]},"delivery":{"work":"Measure the backlog, set a deletion schedule, build a retention filter and repeat the impact assessment.","dependencies":["Case outcome data from courts","Records system vendor support","Legal sign-off on retention rules"],"ownership":"Records manager and data protection officer, with chief officer sign-off.","skillsAdoption":"Records staff need procedures for outcome-based deletion; officers need clear rules on what tools are off until the gate is passed.","governanceCheckpoints":["Backlog measured","Impact assessment repeated","Chief officer approval before pilot"],"acceptanceCriteria":["Overdue-image count reported and falling on schedule","Retention filter verified on a test set","No biometric tool live until the assessment is passed"],"risks":["Backlog too large to clear","Pressure to launch before clean-up","National data outside force control"]}},"retrievedAt":"2026-11-02T14:34:05Z","enrichedAt":"2026-11-02T14:41:23Z","enrichmentBasis":"retrieved source","accessibilityWorkforceImplications":"Clearing a 300,000-person backlog would take records staff time; the Home Office has called mass review resource-intensive.","procurementImplications":"Biometric tool purchases should be conditioned on a lawful, cleaned gallery, not just on the vendor’s matching accuracy.","operatingModelImplications":"Retention review is a standing records-management duty that must run before and alongside any biometric search service.","sourceVerification":{"openedUrl":"https://www.liverpoolecho.co.uk/news/liverpool-news/merseyside-police-unlawfully-holds-300000-34834901","referenceExcerpt":"There is a real risk the force could use and compare against an unlawful image whilst undertaking an OIFR.","promptVersion":"sled-research-v4.0","model":null,"basis":"agent-reported inspection"}}}]}