{"resourceId":"nevada-plain-language-nascio-taxation-2026","versions":[{"version":"external-7c590a9478e1a5ae0a6bde8c9bc5125c01cacbda7e71cc4d7555dfc7ac2592a2","resource":{"id":"nevada-plain-language-nascio-taxation-2026","title":"Nevada Taxation NASCIO: AI plain-language notices cut grade level 13.4→8.4 and general-question calls ~52%","organization":"Nevada Department of Taxation (NASCIO State IT Recognition Awards nomination)","sector":"State tax administration / citizen correspondence / generative AI drafting","geography":"Nevada statewide Department of Taxation; model noted for Secretary of State Commercial Recordings plain-language work","publishedAt":"2026-08-01","publicationDate":"2026-08-01","eventDate":"2025-12-31","sourceName":"NASCIO / Nevada Department of Taxation","sourceLabel":"Plain Language at Scale: Responsible AI for Taxpayer Correspondence","sourceUrl":"https://www.nascio.org/wp-content/uploads/2026/08/NV_ArtificialIntelligence.pdf","evidenceClass":"public-sector-association","outcomeClass":"effective","topics":["knowledge-work","accessibility-workforce","governance-procurement","operating-model"],"finding":"Nevada Taxation’s NASCIO Artificial Intelligence nomination documents a governed GenAI drafting workflow for taxpayer notices (project dates Oct 1, 2024–Nov 30, 2026; operational releases underway). Staff generate first drafts with enterprise-approved AI from public-domain templates—no taxpayer data in prompts—then Quality Forms Committee / business owners review for legal meaning. Reported results: Flesch-Kincaid grade level 13.4→8.4; Reading Ease 33.9→52.7; 119 letters in production covering 67% of system letter volume and 65% of unique taxpayers; general-question calls 23,121 (2023)→11,175 (2025).","sledRelevance":"Strong measured citizen-communication GenAI outcome left unarchived after Sep 30 award-week coverage of Caltrans/Ohio. Why now: primary nomination PDF still not in archive; supplies readability + call-volume metrics transferable to state contact-center and notice modernization after NASCIO awards.","evidence":"Opened the NASCIO HTTPS PDF via curl (UA SLED-research-bot/1.0) and pdftotext; confirmed metrics table (grade level, Reading Ease, 119 letters, call volumes), governance model (human approval, no taxpayer data in prompts), and Secretary of State reuse note.","architectureImplications":"Pattern is prompt+scoring+human release over high-volume templates—not a citizen chatbot or agent. Keep AI out of case PII; wire readability scoring into CMS/letter production; retain committee release gates.","governanceImplications":"Nomination argues AI created speed while governance created trust—Quality Forms Committee ownership is the control, not model choice alone.","securityPrivacyImplications":"Explicitly avoids putting taxpayer data in prompts; still requires secure enterprise AI tenancy and audit of prompt libraries.","caveats":"Self-reported awards nomination; call-volume drop is directional and may include category noise; not an independent audit; PDF upload dated in Aug 2026 path though metrics cover 2023–2025 operations.","streamIds":["state-government"],"roles":{"sales":"Frame Nevada’s Taxation nomination as a measured plain-language GenAI pattern—grade-level and ~52% general-question call cuts under human release—not a chatbot pitch. Discovery should ask which high-volume notices have readability baselines, whether taxpayer data is barred from prompts, and who owns the quality-forms gate. Bounded engagement: a 60–90 day plain-language AI pilot on a named letter family with before/after Flesch scores and call-category tracking. Unsupported claims include guaranteed 52% call reduction, that any GenAI vendor is required, or that awards nominations equal independent audits.","engineering":"Fit is AI-assisted first-draft rewriting of public templates with objective readability scoring and mandatory human legal/comms approval—not autonomous notice generation to residents. Architecture: enterprise GenAI tenant to template library to scoring service to Quality Forms workflow to CMS release, with no case PII in prompts. Prerequisites: template inventory, readability baseline, committee SOP, approved AI tenancy. Constraints: self-reported nomination metrics; call taxonomy messiness. Proof of value is paired grade-level/Reading Ease deltas and sampled legal-accuracy acceptance, not call drops alone.","delivery":"Delivery inventories top letter volume, runs AI draft+human review cycles, instruments call categories, and documents release ownership. Dependencies: Taxation IT, legal, communications, contact center, enterprise AI platform. Ownership of production letters stays with business/legal owners. Acceptance criteria: ≥N letters released with measured grade-level improvement and zero unauthorized PII-in-prompt findings. Risks: treating call drops as causal without controls, and skipping legal review under speed pressure."},"roleDetails":{"sales":{"customerProblem":"States send high-volume notices residents cannot understand, driving avoidable call volume.","stakeholders":["Tax agency CIO/IT","Legal/counsel","Communications","Contact-center leads","Procurement"],"discoveryQuestions":["Which letters drive general-question calls?","What readability baseline exists?","Is taxpayer data barred from prompts?","Who signs production release?"],"valueHypothesis":"Governed GenAI drafting can improve readability and cut routine calls if human legal gates and baselines are enforced.","potentialEngagement":"60–90 day plain-language AI pilot on one high-volume letter family with scoring and call tracking.","unsupportedClaims":["Guaranteed 52% call reduction","Vendor-specific necessity","Award nomination equals audit","Chatbot substitutes for notice rewrite"]},"engineering":{"fit":"Template first-draft GenAI with readability scoring and human legal release.","architectureIntegration":"Enterprise GenAI to public templates to scoring to Quality Forms Committee to CMS.","prerequisites":["Template inventory","Readability baseline","Committee SOP","Approved AI tenant"],"constraints":["Self-reported metrics","Call-category noise","No independent audit"],"security":"No taxpayer PII in prompts; controlled prompt library; release logging.","proofOfValue":["Grade-level before/after","Legal acceptance rate","Sampled call-category change"]},"delivery":{"work":"Select letter set; run AI+review cycles; publish with metrics; train reviewers.","dependencies":["Agency legal","Comms","Contact center","Enterprise AI","CMS owners"],"ownership":"Business/legal owners release letters; IT owns tooling.","skillsAdoption":"Reviewer checklists for legal meaning vs readability tradeoffs.","governanceCheckpoints":["Baseline lock","Pilot mid-review","Production release approval"],"acceptanceCriteria":["Measured readability gain on released letters","No PII-in-prompt incidents","Documented owner sign-off"],"risks":["Causal overclaim on calls","Legal meaning drift","Bypassing committee"]}},"retrievedAt":"2026-10-01T22:23:40-05:00","enrichedAt":"2026-10-01T22:23:40-05:00","enrichmentBasis":"retrieved source","accessibilityWorkforceImplications":"Five-grade readability gain and multilingual translation path improve access for taxpayers; staff shift from drafting burden to legal/comms review skills.","procurementImplications":"Reusable with common office tools and public templates—procurement should buy governed enterprise GenAI access and scoring, not bespoke notice-AI SaaS without human release rights.","operatingModelImplications":"Scale plain-language AI only where a standing content governance board owns baselines, scoring, and production release—not as one-off marketing rewrites.","sourceVerification":{"responseId":"curl-nv-plain-language-20261001T222340","openedUrl":"https://www.nascio.org/wp-content/uploads/2026/08/NV_ArtificialIntelligence.pdf","referenceExcerpt":"Average Flesch-Kincaid grade level dropped from 13.4 to 8.4","promptVersion":"sled-research-v4.0","model":null,"basis":"agent-reported inspection"}}}]}