{"resourceId":"fjp-ai-police-reports-risks-brief-202506","versions":[{"version":"external-8a13217d643ab286ff7c9f005bf8721fb3f56e5ae44ee9251c7c5b63fd33aa71","resource":{"id":"fjp-ai-police-reports-risks-brief-202506","title":"Fair and Just Prosecution: AI-generated police reports pose accuracy, accountability, and Brady risks","organization":"Fair and Just Prosecution","sector":"Prosecution / criminal justice policy","geography":"United States (national prosecutor network; King County, WA case study)","publishedAt":"June 2025","publicationDate":"2025-06-01","eventDate":null,"sourceName":"Fair and Just Prosecution","sourceLabel":"Prosecutor-network issues brief","sourceUrl":"https://fairandjustprosecution.org/wp-content/uploads/2025/06/AI-Generated-Police-Reports-High-Tech-Low-Accuracy-Big-Risks-June-2025.pdf","evidenceClass":"public-sector-association","outcomeClass":"cautionary","topics":["knowledge-work","governance-procurement","data-security","operating-model"],"finding":"FJP’s June 2025 Issues at a Glance brief argues AI-generated police reports present high-stakes risks: hallucination/inaccuracy from poor audio and overlapping speakers; cognitive offloading that dilutes officer ownership; bias amplification; privacy exposure via vendor processing of BWC audio; minimal proven time savings (citing Adams et al.); and legal exposure including Fourth Amendment fact errors, Brady demands for model details, and impeachment if officers certify flawed AI text. Case study: King County Prosecuting Attorney’s Office does not accept AI-assisted narratives and requires reports produced entirely by the authoring officer after seeing an AI report reference an officer not at the scene. Recommendations urge prosecutors to set non-acceptance or strict safeguard policies and demand independent validation before adoption. Why now: newly relevant to 2026-09 Michigan Draft One adoption and Wayne County prosecutor caution documented in this edition’s WXYZ source.","sledRelevance":"Independent prosecutor-side scrutiny framework that explains why DA acceptance gates—not vendor demos—should control AI report rollout in SLED agencies now piloting Draft One.","evidence":"Opened FJP PDF via WebFetch and curl/pdftotext; confirmed King County non-acceptance case study, risk categories, Adams time-savings citation, and recommendation that prosecutors decline or tightly govern AI reports.","architectureImplications":"If used at all, isolate vendor LLM drafting from final RMS narrative authority; preserve prompts/versions for Brady; prefer prosecutor-reject flags in RMS.","governanceImplications":"Prosecutors should publish acceptance/non-acceptance rules before PD purchase; require independent accuracy studies and disclosure protocols.","securityPrivacyImplications":"BWC audio to vendors creates retention and misuse risks; contracts need CJIS, deletion, and no-train clauses.","caveats":"Policy brief (June 2025), not a new RCT; King County example is office policy anecdote; quantitative claims largely cite external studies; why-now depends on current Michigan adoption context rather than a 2025–26 update to the PDF itself.","streamIds":["public-safety"],"roles":{"sales":"Sales framing: FJP’s cautionary prosecutor brief is the scrutiny counterpart to Michigan Draft One pilots—arguing AI report tools can jeopardize prosecutions via hallucinations, diluted accountability, bias, privacy leakage, and Brady fights, and highlighting King County’s non-acceptance policy. Engage DAs and chiefs on acceptance rules before sell-side efficiency pitches. Bounded engagement: prosecutor policy workshop and independent validation plan. Unsupported: guaranteed time savings or that Easter-egg tests fully mitigate courtroom risk.","engineering":"Engineering framing: Design assumes AI drafts are untrusted until officer-authored final narrative standards are met; support non-acceptance modes and discovery of AI involvement. Prerequisites: contract transparency for models/prompts, audio quality controls, and RMS labeling. Constraints: proprietary models hinder Brady production. Security: minimize BWC audio egress and ban vendor training reuse. Proof of value: independent accuracy trials and zero AI-only narratives accepted for charging.","delivery":"Delivery framing: Help prosecutor and PD leaders implement a written acceptance or non-acceptance policy, train officers on cognitive-offloading and attestation risks, and stand up audit sampling of AI-touched reports. Dependencies include DA leadership, PD vendor contracts, and defense discovery expectations. Ownership: elected prosecutor for charging standards; chief for field compliance. Acceptance: published DA rule in force and AI assistance detectable in case files. Risks include impeachment cascades and public-trust loss from certified AI errors."},"roleDetails":{"sales":{"customerProblem":"Prosecutors need charging-ready police narratives while vendors push AI drafting that may introduce unreviewable errors and discovery burdens.","stakeholders":["District/State’s Attorney","Police chief","Defense bar","Court clerks","Vendor account team"],"discoveryQuestions":["Does the DA accept AI-assisted narratives today?","What Brady materials can the vendor produce?","Have independent accuracy studies been run on local audio?","How is AI assistance disclosed in discovery?"],"valueHypothesis":"Bounded value is clearer acceptance rules and safer pilots—not proven efficiency or accuracy gains.","potentialEngagement":"Prosecutor–PD joint policy sprint defining non-acceptance or safeguard conditions before any expansion.","unsupportedClaims":["Vendor-claimed time savings as proven","That officer review eliminates hallucination risk","Universal fitness of Draft One for all case types"]},"engineering":{"fit":"Only as tightly controlled draft assist—or disabled—when prosecutor policy forbids AI-assisted narratives.","architectureIntegration":"Optional draft service behind policy flag; force officer rewrite path; persist AI-use metadata for Brady; block silent AI finalize.","prerequisites":["DA written policy","Vendor disclosure pack","RMS AI-flag","Audio quality thresholds"],"constraints":["Proprietary model opacity","Brief is not a local empirical study","Overlapping-speech/noise failure modes"],"security":"CJIS-aligned vendor controls; deletion SLAs; no secondary use of BWC audio for model training.","proofOfValue":["Independent accuracy evaluation","Prosecutor challenge rate","Impeachment incidents tied to AI text = 0"]},"delivery":{"work":"Draft DA acceptance policy, train officers on ownership/attestation, and stand up audit sampling of AI-touched reports.","dependencies":["FJP-style policy models","PD vendor contracts","Discovery unit procedures"],"ownership":"Elected prosecutor owns acceptance standard; chief owns field compliance.","skillsAdoption":"Reinforce that certifying a report means personal knowledge—not rubber-stamping AI prose.","governanceCheckpoints":["Policy issuance","First-month audit","Revisit after any AI-error incident"],"acceptanceCriteria":["Published DA rule on AI reports","AI assistance detectable in case files","Sampling finds no unedited Easter-egg/hallucinated content"],"risks":["Brady litigation over model secrecy","Officer Brady-list exposure","Erosion of public trust after AI errors"]}},"retrievedAt":"2026-09-20T03:10:34Z","enrichedAt":"2026-09-20T03:13:30Z","enrichmentBasis":"retrieved source","accessibilityWorkforceImplications":"Offices without training capacity for automation-bias resistance should not expand AI drafting; FJP stresses skill and accountability erosion risks.","procurementImplications":"Require independent evaluations, Brady-producible documentation, and DA acceptance clauses before AI report tool purchase.","operatingModelImplications":"Make prosecutor acceptance the control point for AI police-report operating models.","sourceVerification":{"openedUrl":"https://fairandjustprosecution.org/wp-content/uploads/2025/06/AI-Generated-Police-Reports-High-Tech-Low-Accuracy-Big-Risks-June-2025.pdf","referenceExcerpt":"must be produced entirely by the authoring officer","promptVersion":"sled-research-v4.0","model":null,"basis":"agent-reported inspection"}}}]}