{"resourceId":"salt-lake-county-da-ai-evidence-review-closed-systems-disclosure-20261023","versions":[{"version":"external-38ae6ef545ea9c39acda6965bd19bbf3bfcd3f4c67aa824efb1481f75d0b44cf","resource":{"id":"salt-lake-county-da-ai-evidence-review-closed-systems-disclosure-20261023","title":"Salt Lake County and Montgomery County, Texas, prosecutors describe AI for evidence review and legal research, and ask police to disclose AI-drafted reports","organization":"Salt Lake County District Attorney’s Office; Montgomery County (Texas) District Attorney’s Office","sector":"Prosecution / AI-assisted evidence review","geography":"Utah and Texas, United States","publishedAt":"Oct. 23, 2026","publicationDate":"2026-10-23","eventDate":null,"sourceName":"KSL.com","sourceLabel":"Salt Lake County prosecutors using AI to review evidence; DA says humans remain in control","sourceUrl":"https://www.ksl.com/article/51640364/salt-lake-county-prosecutors-using-ai-to-review-evidence-da-says-humans-remain-in-control","evidenceClass":"independent-reporting","outcomeClass":"emerging","topics":["knowledge-work","data-security","operating-model"],"finding":"Speaking around the 2026 National Prosecution Best Practices Conference in Salt Lake City, District Attorney Sim Gill said his office uses AI to organise body-camera footage into searchable transcripts and to point attorneys to relevant moments, and uses AI legal research tools inside closed systems; it does not use AI to write motions or make charging decisions. He said he has told police that officers using AI-drafted reports should disclose it and verify the content. Montgomery County, Texas, DA Mike Holley described using an AI tool to isolate relevant Snapchat messages in a child sexual abuse case, with prosecutors checking flagged messages against originals, and a tool that compares warrant affidavits to original reports. Both cited hallucination, data security and deskilling risks; Gill said efficiency does not automatically lower costs.","sledRelevance":"Shows how prosecutors are setting their own conditions for AI use, including a disclosure expectation for AI police reports that affects agencies using Draft One and similar tools.","evidence":"Opened the KSL article and read it in full.","architectureImplications":"Closed, access-controlled AI research and evidence tools are preferred over public platforms; transcripts and flags must link back to original evidence.","governanceImplications":"Gill describes expectations for disclosure and verification but, per the article, no newly adopted written policy on AI police reports.","securityPrivacyImplications":"Both DAs name control over where uploaded evidence goes, how it is stored and who can retrieve it as a primary concern.","caveats":"Interview-based reporting; no time savings, error rates or tool names are measured or verified. Statements describe intentions and practice in two offices.","streamIds":["public-safety"],"roles":{"sales":"Sales framing: prosecutors are open to AI for evidence triage and research but set clear limits on drafting and decisions. The customer problem is digital evidence volume against thin staffing. A bounded engagement could assess one evidence-review workflow, such as body-camera triage, with verification steps. This evidence cannot support claims of time or cost savings; Gill explicitly cautions that efficiency may not lower costs, and no measurements are reported.","engineering":"Engineering framing: the requirements here are closed deployment, traceability from every AI flag back to the source evidence, and clarity on data storage and retrieval. Pre-sales should check integration with evidence management systems and whether transcripts carry timestamps to original footage. A proof of value could compare AI-flagged moments with attorney review on a closed case set, recording misses and false flags, since the article offers no accuracy data.","delivery":"Delivery framing: delivery would define which tasks AI may assist (triage, transcription, research) and which it may not (motions, charging), with verification recorded before evidence is used. Dependencies include evidence platform access and police agreement on disclosing AI-drafted reports. The operational owner is the DA’s office. Acceptance should be measured by verified flags and documented review, and risks include over-reliance and skill loss among newer attorneys."},"roleDetails":{"sales":{"customerProblem":"Prosecutors face growing digital evidence volume with limited staff and need help without surrendering judgment or exposing sensitive data.","stakeholders":["District attorneys","Trial and appellate attorneys","Investigators and paralegals","Office IT and security","Partner police agencies"],"discoveryQuestions":["Which evidence types consume the most review time?","Where would uploaded evidence be stored, and who could retrieve it?","Do partner agencies disclose when reports were drafted with AI?"],"valueHypothesis":"AI triage of body-camera and message evidence could direct attorney attention faster, provided every flag is verified against the original.","potentialEngagement":"Pilot assessment of one evidence-review workflow on closed cases, with a verification log and data-handling review.","unsupportedClaims":["Measured time or cost savings","That AI research tools are hallucination-free","That a disclosure policy for AI police reports has been adopted"]},"engineering":{"fit":"Prosecutor offices with heavy body-camera and digital message evidence.","architectureIntegration":"Integrates with evidence management to transcribe and index footage; flags link to timestamps in original files; research tools run in closed environments.","prerequisites":["Evidence platform integration","Data residency and retention terms","Attorney verification workflow"],"constraints":["No accuracy data in the source","Discovery and disclosure obligations vary by state","Closed systems may limit model choice"],"security":"Keep evidence inside closed, access-controlled systems with audit logs; prohibit public-platform uploads.","proofOfValue":["Proposed: compare AI-flagged moments with full attorney review on closed cases and record misses","Proposed: confirm each transcript segment links back to original footage"]},"delivery":{"work":"Define permitted AI tasks, set verification steps and agree disclosure expectations with partner police agencies.","dependencies":["Evidence platform vendor","Police agency cooperation","Training time for attorneys"],"ownership":"District attorney’s office","skillsAdoption":"Train attorneys to verify AI output against source evidence and to maintain core research and interviewing skills.","governanceCheckpoints":["Written list of permitted and prohibited AI uses","Verification log review","Annual data-handling review"],"acceptanceCriteria":["Proposed: every AI flag used in court is verified and logged","Proposed: written disclosure protocol agreed with partner agencies"],"risks":["Hallucinated research summaries","Sensitive evidence leaving controlled systems","Deskilling of junior attorneys"]}},"retrievedAt":"2026-10-25T02:57:20Z","enrichedAt":"2026-10-25T03:02:27Z","enrichmentBasis":"retrieved source","accessibilityWorkforceImplications":"Staffing shortages drive interest; both warn about junior attorneys losing research and interviewing skills.","procurementImplications":"Interpretation: prosecutor offices will ask vendors about data residency, retention and access before buying AI review tools.","operatingModelImplications":"Human review of every AI flag before use in court; AI kept out of charging, plea and diversion decisions.","sourceVerification":{"openedUrl":"https://www.ksl.com/article/51640364/salt-lake-county-prosecutors-using-ai-to-review-evidence-da-says-humans-remain-in-control","referenceExcerpt":"Gill said his office also uses AI-assisted legal research tools within closed systems rather than open, publicly accessible platforms","promptVersion":"sled-research-v4.0","model":null,"basis":"agent-reported inspection"}}}]}