{"resourceId":"florida-sboe-6a-10957-ai-change-notice-20260918","versions":[{"version":"external-b1ff4488bd54c3289493bdde6da036eb3c000c0abb84f540d2f38c34c6fc1acb","resource":{"id":"florida-sboe-6a-10957-ai-change-notice-20260918","title":"Florida FAR Change notice (Sept 18): AI never decision-maker without human review; agentic tools need kill-switch","organization":"Florida State Board of Education / Florida Department of Education (Rule 6A-1.0957)","sector":"K-12 state board administrative rulemaking","geography":"Florida, United States (statewide public schools and charters)","publishedAt":"Notice of Change published September 18, 2026, Florida Administrative Register Vol. 52/182 (amending Aug 26 proposed rule after Sept 16 hearing)","publicationDate":"2026-09-18","eventDate":"2026-09-16","sourceName":"FAR Notice of Change 31384852 — Rule 6A-1.0957 Internet Safety Policy (official notice file)","sourceLabel":"State Board of Education notice of change to proposed AI internet-safety rule","sourceUrl":"https://flrules.org/gateway/notice_Files.asp?ID=31384852","evidenceClass":"standards-guidance","outcomeClass":"emerging","topics":["governance-procurement","data-security","accessibility-workforce","operating-model"],"finding":"On September 18, 2026, Florida published a Notice of Change to Rule 6A-1.0957 clarifying post-hearing amendments: definitions expanded to generative and autonomous/agentic AI instructional tools; districts/charters must amend internet-safety policies by July 1, 2027. The change requires that agentic/autonomous AI in VPK–12 be used only under direct adult supervision, maintain complete activity logs, and include a mechanism for school personnel to immediately suspend/disable autonomous functionality without vendor assistance. It also requires that AI never be used as a decision maker absent human review for grading, evaluations, promotions, discipline, threat assessment, counseling/ESE referrals, or communications, and that districts disclose unexpected autonomous behavior or IT safety/security incidents from approved tools to the Department. Broader proposed package (opened companion notice) still includes parent notification/opt-in with substantively comparable non-AI alternatives, PreK–5 developmental review, bans on companion/anthropomorphic relationship design, and staff training.","sledRelevance":"A binding statewide operating floor for student AI tools—parent voice, companion bans, human-in-the-loop decision bans, and vendor kill-switches—that other SEAs/LEAs will copy or contrast with guidance-only models.","evidence":"Opened official FAR Change notice DOC (ID 31384852) via flrules notice_Files and extracted full amendment text; cross-checked View_notice HTML (publish date 9/18/2026) and the Aug 26 proposed-rule notice DOC (ID 31301432) for parent opt-in/companion/training provisions; noted same-day WCTV coverage but cited the official notice as verification object.","architectureImplications":"Vendors must ship adult-supervised agentic modes, complete activity logs, and district-controlled immediate disable switches independent of vendor support; high-stakes decision features need hard human-review gates.","governanceImplications":"LEAs must rewrite internet-safety policies by July 1, 2027 with public approved-tool lists, parent opt-in workflows, and DOE incident disclosure paths—not optional guidance.","securityPrivacyImplications":"Incident disclosure for unexpected autonomous behavior and unauthorized data access becomes a compliance reporting duty; activity logs create retention/access obligations for parents and districts.","caveats":"Change notice amends proposed text; final FAC codification timing may lag; higher-ed portions of political messaging are out of k12 stream scope here—we cite K-12 instructional-tool provisions only. Tallahassee Democrat Sept 16 coverage already archived—this edition uses the Sept 18 official Change notice as the fresh primary.","streamIds":["k12"],"roles":{"sales":"Sales framing: Florida just clarified a binding kill-switch and no-AI-as-decision-maker floor. Lead with adult-supervised agentic modes, parent opt-in alternatives, and DOE incident disclosure—not companion chatbots. Ask whether your product can be disabled by school staff without a vendor ticket and whether grading/discipline AI can be forced human-review-only. Position a July 2027 policy-compliance readiness engagement that inventories tools against 6A-1.0957 controls.","engineering":"Engineering framing: implement immediate autonomous-disable controls, immutable activity logs, and feature flags that block AI decisioning for grades, discipline, threat assessment, and ESE referrals. Parent opt-in needs roster-aware entitlement switching to substantively comparable non-AI paths. Prove PreK–5 content review hooks and no anthropomorphic companion behaviors in student builds, and ship DOE-ready incident export for unexpected autonomous behavior.","delivery":"Delivery framing: map current AI tools to 6A-1.0957 controls, draft internet-safety policy amendments, stand up parent notice/opt-in, train staff on risks/limitations, and tabletop kill-switch and incident disclosure. Acceptance is a published approved-tool list, working disable mechanism, and documented human-review gates on high-stakes decisions before July 1, 2027. Keep charter boards on the same amendment calendar."},"roleDetails":{"sales":{"customerProblem":"Florida districts must amend AI internet-safety policies by July 1, 2027 under binding rules including parent opt-in and agentic kill-switches.","stakeholders":["Superintendents","School boards","CTOs","General counsel","Principals","Parent leaders"],"discoveryQuestions":["Can staff disable autonomous AI features without vendor assistance?","Where does AI touch grading, discipline, or ESE referrals today?","Is parent opt-in with comparable non-AI alternatives operational?","How will you disclose unexpected autonomous behavior to DOE?"],"valueHypothesis":"Vendors that ship kill-switches, activity logs, and human-review gates will clear Florida (and copycat SEA) compliance faster than companion-style student AI.","potentialEngagement":"4–6 week 6A-1.0957 gap assessment: tool inventory, policy draft, parent-notice workflow, and kill-switch tabletop.","unsupportedClaims":["That compliance alone improves learning outcomes","Guaranteed immunity from parent opt-out operational cost","That guidance-only states have identical mandates"]},"engineering":{"fit":"Fit for enterprise instructional AI with admin controls; poor fit for unsupervised companion or agentic student products.","architectureIntegration":"Roster entitlements for opt-in → supervised agent runtime → activity log store → district kill-switch API → human-review workflow for high-stakes outputs.","prerequisites":["Student information system roster hooks","Admin disable API","Activity logging retention ≥30 days where required","Incident reporting channel to SEA/DOE"],"constraints":["July 1, 2027 local policy deadline","PreK–5 extra developmental review","No AI decision-maker without human review"],"security":"Log and report unexpected autonomous behavior and unauthorized data access; keep student interaction records available for parental FERPA access.","proofOfValue":["Demonstrate kill-switch without vendor ticket","Show blocked AI grading/discipline decision path","Parent opt-out routes to comparable non-AI tool","Sample DOE incident disclosure packet"]},"delivery":{"work":"Inventory tools, amend internet-safety policy, configure opt-in and kill-switch, train staff, publish approved-tool list, and schedule compliance audit.","dependencies":["Legal","IT security","Curriculum","Communications","Charter governing boards"],"ownership":"Superintendent and board own policy adoption; CTO owns technical controls; principals own supervised use.","skillsAdoption":"Mandatory teacher/admin training on AI risks, limitations, safety, and rule obligations before tool use.","governanceCheckpoints":["Board adoption of amended policy","Public approved-tool list live","Kill-switch drill completed","Pre–July 2027 compliance attestation"],"acceptanceCriteria":["Policy text covers opt-in, companions ban, human-review bans","Disable mechanism tested in production","No AI auto-decision on listed high-stakes uses","Incident disclosure SOP approved"],"risks":["Vendor cannot provide kill-switch","Shadow companion tools persist","Opt-out alternative quality disputes"]}},"retrievedAt":"2026-09-20T03:05:20Z","enrichedAt":"2026-09-20T03:10:38Z","enrichmentBasis":"retrieved source","accessibilityWorkforceImplications":"Staff training is mandatory under the proposed package; supervised agentic use increases adult oversight load—plan coverage for aides/subs and preserve IEP assistive exceptions carefully.","procurementImplications":"Contracts must require vendor-independent disablement, activity logs, no companion/anthropomorphic student design, and cooperation on DOE incident disclosure.","operatingModelImplications":"Internet-safety policy becomes the AI operating system: annual tool list updates, parent notices, and kill-switch drills—not ad hoc principal permission.","sourceVerification":{"openedUrl":"https://flrules.org/gateway/notice_Files.asp?ID=31384852","referenceExcerpt":"Ensure that AI is never used as a decision maker absent human review for any use case","promptVersion":"sled-research-v4.0","model":null,"basis":"agent-reported inspection"}}}]}