{"resourceId":"ncsc-jtc-ai-assisted-transcription-courts-governance-safeguards-202609","versions":[{"version":"external-3b71819e9fb9b5041a701366fc1201336b5bfef9e610f75dfb3c6a61aa6d6215","resource":{"id":"ncsc-jtc-ai-assisted-transcription-courts-governance-safeguards-202609","title":"Court technology committee guidance: AI transcripts should stay unofficial drafts until reviewed, and courts should validate on representative local recordings","organization":"Joint Technology Committee (COSCA, NACM and National Center for State Courts)","sector":"State courts (public-sector association guidance)","geography":"United States","publishedAt":"2026-09 (version 1.0 approved 2026-09-18)","publicationDate":"2026-09-18","eventDate":null,"sourceName":"National Center for State Courts","sourceLabel":"AI-Assisted Transcription in Courts: Governance, Use, and Safeguards (JTC Resource Bulletin)","sourceUrl":"https://ncsc.contentdm.oclc.org/digital/api/collection/tech/id/1307/download","evidenceClass":"public-sector-association","outcomeClass":"cautionary","topics":["governance-procurement","accessibility-workforce","data-security","knowledge-work"],"finding":"Drawing on a national scan of state courts, the bulletin says AI transcription accuracy may vary materially with accents, dialects, overlapping speech, interpreter-mediated exchanges, recording quality and acoustics. Courts should validate on representative local recordings rather than vendor claims or generic metrics, set documented quality thresholds, and require human review before AI text is used for filing, certification or appeals. The most common emerging model treats AI text as an unofficial draft, with heightened safeguards for confidential, sealed and juvenile matters and attention to retention, discoverability and vendor data use.","sledRelevance":"Newly relevant because NVIDIA's Oct 28 release makes a self-hostable, speaker-attributing transcription model freely available; this is the governing U.S. court-administration guidance any state or local court would apply before piloting it, and its validation criteria transfer to clerks and public meetings.","evidence":"Public-sector association guidance based on a national scan of state courts; governance recommendations, not an accuracy study.","architectureImplications":"Pilots should log tool, version and date with each transcript and keep AI drafts distinct from the certified record.","governanceImplications":"Requires defined legal status for AI text, review and certification procedures, and possibly rulemaking before official use.","securityPrivacyImplications":"Calls for confidentiality, retention, discoverability and vendor data-use controls, with heightened safeguards for sealed and juvenile proceedings; self-hosting can help but does not settle legal status.","caveats":"Guidance, not empirical measurement; does not evaluate NVIDIA or any specific model; written for courts, so other agencies should adapt it.","streamIds":["nvidia"],"roles":{"sales":"For court and clerk buyers, this bulletin defines what a responsible offer must include. Lead with governance: validation on the court's own recordings, documented quality thresholds and human review before official use. Pairing an open model like Nemotron Transcribe with that framework is a credible pitch, especially where reporter shortages delay transcripts. Never position AI output as a certified transcript. A pilot scoped to draft transcripts and searchable recordings fits the bulletin's most common emerging model.","engineering":"Engineering should build pilots to produce the evidence the bulletin asks for: word-level accuracy, material-error rates, correction burden, turnaround and speaker identification on representative local recordings, including interpreters and poor audio. Tag every output as an unofficial draft with tool, version and timestamp metadata. Keep sealed and juvenile proceedings out of scope or under separate controls. Self-hosting helps with vendor data-use concerns but does not change the transcript's legal status.","delivery":"Delivery should involve court leadership, the clerk and counsel before any pilot to define the legal status of AI text and who reviews it. Write quality thresholds and review procedures into the pilot charter. Exclude confidential, sealed and juvenile matters at first. Track correction burden and material errors, and report them to the court's technology governance body. Plan for rule or policy changes if AI text is ever to support filings or appeals."},"roleDetails":{"sales":{"customerProblem":"Courts face transcript delays and reporter shortages but cannot risk the integrity of the record.","stakeholders":["Court administrator","Chief judge or presiding judge","Clerk of court","Court technology committee","General counsel"],"discoveryQuestions":["What legal status would AI text have in your court?","Which proceedings are excluded (sealed, juvenile)?","Who reviews and certifies transcripts today?"],"valueHypothesis":"A governed draft-transcript pilot could shorten access to hearing records while preserving certification.","potentialEngagement":"Draft-transcript pilot on non-confidential proceedings with documented validation.","unsupportedClaims":["Certified or official transcripts from AI alone","Replacement of court reporters","Accuracy claims without local validation"]},"engineering":{"fit":"Draft transcripts and searchable recordings for non-confidential proceedings.","architectureIntegration":"Integrate with digital recording systems; tag drafts with tool, version and timestamp.","prerequisites":["Representative local courtroom recordings","Human reference transcripts","Metadata and audit logging"],"constraints":["AI text generally unofficial until reviewed","Heightened safeguards for sealed and juvenile matters"],"security":"Control retention, discoverability and access; avoid vendor training on court audio.","proofOfValue":["Material-error rate on local recordings","Correction burden per hour","Speaker-identification accuracy"]},"delivery":{"work":"Charter a governed pilot with defined legal status, review and quality thresholds.","dependencies":["Court leadership approval","Counsel review","Recording system access"],"ownership":"Court administration with clerk ownership of the record.","skillsAdoption":"Train reviewers and staff on draft status and correction procedures.","governanceCheckpoints":["Legal status defined before pilot","Exclusion of confidential matters","Governance body review of results"],"acceptanceCriteria":["Proposed: material-error rate below documented threshold","Proposed: reduced turnaround for draft transcripts"],"risks":["Draft mistaken for official record","Confidential audio exposure","Unequal accuracy for interpreted or accented speech"]}},"retrievedAt":"2026-10-29T02:59:42Z","enrichedAt":"2026-10-29T03:06:28Z","enrichmentBasis":"retrieved source","accessibilityWorkforceImplications":"Notes court reporter shortages and accessibility benefits, while keeping certified reporting and human reviewers central.","procurementImplications":"Recommends procurement terms on data usage, security, auditability and confidentiality, and caution about vendor-reported accuracy.","operatingModelImplications":"Courts need review staffing, quality monitoring and documentation of AI use in each transcript workflow.","sourceVerification":{"openedUrl":"https://ncsc.contentdm.oclc.org/digital/api/collection/tech/id/1307/download","referenceExcerpt":"Courts should test system performance under representative local courtroom","promptVersion":"sled-research-v4.0","model":null,"basis":"agent-reported inspection"}}}]}