{"resourceId":"uscourts-ai-interim-guidance-20260917","versions":[{"version":"external-8a13217d643ab286ff7c9f005bf8721fb3f56e5ae44ee9251c7c5b63fd33aa71","resource":{"id":"uscourts-ai-interim-guidance-20260917","title":"U.S. Courts: Judicial Conference update on AI task force and interim non-delegation guidance","organization":"Administrative Office of the U.S. Courts","sector":"Courts / judiciary","geography":"United States (federal judiciary)","publishedAt":"September 17, 2026","publicationDate":"2026-09-17","eventDate":"2026-09-17","sourceName":"uscourts.gov","sourceLabel":"Official federal judiciary news release","sourceUrl":"https://www.uscourts.gov/data-news/judiciary-news/2026/09/17/judiciary-cites-progress-case-management-property-authority-and-ai","evidenceClass":"standards-guidance","outcomeClass":"emerging","topics":["governance-procurement","knowledge-work","operating-model","data-security"],"finding":"On 2026-09-17 the AO reported to the Judicial Conference that the AI advisory task force (appointed 2025) identified more than 60 issues, prioritized them, and formed seven subject-matter subgroups. Director Conrad stated he issued Task Force-developed interim guidance: courts cautioned not to delegate core judicial functions to AI, including decision-making or case adjudication, and all Judiciary users reminded they are accountable for work performed with AI assistance. Additional guidance will roll out as issues are resolved. Ninth Circuit Judge Sidney R. Thomas (task force chair) updated the Conference. Release also covers CM/ECF modernization and real-property authority—not used as AI evidence here.","sledRelevance":"Primary-source federal courts AI control statement useful for state/local court peers; complements prior Reuters secondary coverage already archived by supplying the official interim non-delegation and accountability language.","evidence":"Opened uscourts.gov HTTPS page via WebFetch and curl; confirmed >60 issues, seven subgroups, interim guidance quotes on non-delegation of core functions and user accountability, and Thomas update.","architectureImplications":"Keep generative AI out of adjudication decision paths; confine tools to administrative assists with human sign-off and audit logs.","governanceImplications":"Adopt written non-delegation rules and user accountability attestations before expanding chambers or clerk-office AI.","securityPrivacyImplications":"Pair AI use with IT security program updates (release also stresses CDM coverage for Judiciary systems).","caveats":"Interim guidance text itself is not published in full on this page; no empirical outcome evaluation of AI tools; federal context may differ from state courts; adjacent CMM/RPA content is out of scope for AI findings.","streamIds":["public-safety"],"roles":{"sales":"Sales framing: The Sept 17 AO release is standards-guidance, not a product RFP—the federal judiciary’s interim line is clear: no delegating core judicial functions to AI, and users remain accountable for AI-assisted work. Position advisory help around policy drafting, training, and administrative-only use cases. Stakeholders: chief judges, clerks, AO/IT, and probation/pretrial leaders. Bounded engagement: interim-policy gap assessment for a state or local court mirroring non-delegation rules. Unsupported: claims that federal interim guidance already solves hallucination or litigant-AI problems.","engineering":"Engineering framing: Architect court AI for administrative assists (formatting, triage of non-decisional tasks) with mandatory human accountability markers; block paths that would auto-draft dispositive orders without judge ownership. Prerequisites: written interim policy, access controls, and logging. Constraints: full interim memo not public on this page; >60 issues still under subgroup review. Security: align with Judiciary IT security expectations. Proof of value: policy compliance audits and zero AI-only adjudicative outputs.","delivery":"Delivery framing: Deliver policy communication, staff training on verification of AI-assisted work product, an approved-tool list, and periodic compliance checks for clerk and chambers workflows. Dependencies include further AO/task-force guidance tranches and local IT controls. Ownership sits with the chief judge for policy and the clerk for administrative implementation. Acceptance criteria are documented non-delegation rules plus user accountability attestations on file. Risks include uneven local adoption, hidden AI use, and public-trust damage if AI errors appear in opinions."},"roleDetails":{"sales":{"customerProblem":"Courts need AI productivity aids without surrendering adjudication authority or accountability.","stakeholders":["Chief judge","Clerk of court","Court IT/security","Probation/pretrial","Public defenders (as system users)"],"discoveryQuestions":["Is there a written ban on AI for core judicial functions?","How is user accountability attested and audited?","Which administrative uses are approved now?","How will further task-force guidance be adopted locally?"],"valueHypothesis":"Bounded value is safer administrative AI under non-delegation rules—not AI-decided cases.","potentialEngagement":"2–4 week interim-policy alignment workshop for a state/local court mirroring federal non-delegation language.","unsupportedClaims":["That interim guidance is a complete public rulebook","Measured case-quality improvements from court AI","Resolution of litigant-filed AI hallucinations"]},"engineering":{"fit":"Administrative and drafting assists with human ownership—explicitly not case adjudication engines.","architectureIntegration":"Optional assist tools in clerk/chambers productivity suite; force human approve/sign; deny automated dispositive decision APIs.","prerequisites":["Written interim policy","Role-based access","Audit logging of AI assists","Training curriculum"],"constraints":["Full guidance text not on page","60+ issues still open","Federal vs state procedural differences"],"security":"Keep sensitive case text off unapproved models; align with continuous diagnostics/mitigation expectations cited for Judiciary IT.","proofOfValue":["Compliance sampling of AI-assisted work","No AI-only adjudicative outputs","User accountability attestations on file"]},"delivery":{"work":"Roll out interim policy memo, training, approved-tool list, and periodic compliance checks.","dependencies":["AO/task-force updates","Local IT","Judicial leadership endorsement"],"ownership":"Chief judge owns policy; clerk owns administrative implementation.","skillsAdoption":"Train staff that AI assistance does not transfer responsibility for filed work product.","governanceCheckpoints":["Policy issuance","90-day compliance review","Update when next guidance tranche arrives"],"acceptanceCriteria":["Non-delegation language adopted","Accountability acknowledgment completed by AI users"],"risks":["Uneven local adoption","Hidden AI use in chambers","Public trust hits from AI errors in opinions"]}},"retrievedAt":"2026-09-20T03:10:34Z","enrichedAt":"2026-09-20T03:13:00Z","enrichmentBasis":"retrieved source","accessibilityWorkforceImplications":"Training every Judiciary AI user on accountability is required; capacity varies across courts.","procurementImplications":"Court AI purchases should contractually forbid use for core adjudicative functions and require auditability.","operatingModelImplications":"Separate administrative AI assists from judicial decision-making workflows as a standing operating rule.","sourceVerification":{"openedUrl":"https://www.uscourts.gov/data-news/judiciary-news/2026/09/17/judiciary-cites-progress-case-management-property-authority-and-ai","referenceExcerpt":"courts have been cautioned not to delegate core judicial functions to AI, including decision-making or case adjudication","promptVersion":"sled-research-v4.0","model":null,"basis":"agent-reported inspection"}}}]}