From the K–12 edition of September 17, 2026
Florida Board of Education adopts statewide K-12 AI parent opt-in, companion bans, and vendor kill-switch rules
Florida Board of Education / News Service of Florida via Tallahassee Democrat · K-12 public education policy · Florida, United States
- Publisher
- Tallahassee Democrat (News Service of Florida)
- Original publication
- September 16, 2026
- Source retrieved
- 2026-09-18
- Event date
- 2026-09-16
What happened
Florida moved from scattered building-level AI choices to binding statewide process controls: disclose the tool and purpose, obtain parental consent, provide a comparable non-AI alternative, ban companion/relationship AI and undisclosed profiling, train staff, forbid unsupervised student use, and retain an immediate disable capability without vendor assistance, with district policies due July 1, 2027.
Why it matters
Large-state procurement and internet-safety policy attachment will force edtech vendors serving ~2.9M students to support consent workflows, substitution pathways, U.S. data residency preferences, and operator kill-switches—settings other SLED buyers will copy or be asked to match.
Evidence and measured results
Tallahassee Democrat / News Service of Florida coverage of the Sept 16, 2026 Board of Education meeting quotes Commissioner Henry Mack and details K-12 requirements (parent notice/consent, non-AI alternative of similar instructional quality, PreK-5 extra review, companion AI ban, training, unsupervised-use ban, disable-without-vendor, U.S. data priority, no commercial training on student data). Cross-checked with WFTV (Sept 17) and Florida Press contemporaneous reports.
Limitations and uncertainty
Primary source is news reporting; official FLDOE draft PDF URL returned 404 at retrieval so exact regulatory wording was not verified from the agency file. College-system rules are outside this k12 finding except as statewide context.
Put this evidence to work
Lighthouse Advisory interpretation, grounded in this source. Enriched 2026-09-18; this does not change the original publication date. Each role follows the same checklist so takeaways compare across sources.
Sales
Role takeaway
- Customer problem
- Districts face a binding parent-consent and substitution regime and cannot buy AI tools that lack disclosure, opt-in, and kill-switch controls.
- Stakeholders
- State education agencies
- District superintendents
- Chief technology officers
- School board counsel
- Parent advocacy groups
- Discovery questions
- Can parents see tool name, courses, and interaction mode before consenting?
- What comparable non-AI alternative do you support at similar instructional quality?
- Can the district suspend the product without vendor assistance?
- Where is student data stored, and is commercial model training contractually barred?
- Value hypothesis
- Vendors that ship consent, substitution, and kill-switch controls will clear Florida-style internet-safety procurement faster than feature-rich but uncontrollable tools.
- Potential engagement
- 6–8 week Florida 6A-readiness gap assessment: inventory AI features, map consent/substitution gaps, draft DPA redlines, and produce a July 2027 remediation plan.
- Unsupported claims
- Guaranteed statewide adoption after rule passage
- That opt-in alone improves learning outcomes
- That companion-AI bans eliminate all SEL-related parent concern
Pre-sales engineering
Role takeaway
- Fit
- Fit for districts that can integrate parent portals/SIS with an AI allowlist and runtime policy engine; poor fit for consumer chatbot embeds without admin kill-switch.
- Architecture and integration
- SIS/parent portal consent flags → LMS roster entitlements → secure AI gateway with feature flags; offline/alternate content packs for declined students; admin disable webhook independent of vendor SaaS console.
- Prerequisites
- Authoritative inventory of AI-enabled edtech
- Parent digital consent channel
- Filtering/CASB controls for unsupervised GenAI
- DPA language barring commercial training use
- Constraints
- PreK-5 additional approval path
- No companion/relationship design features
- July 1, 2027 local policy deadline
- Prefer U.S. data residency
- Security
- Enforce FERPA/COPPA-aligned processing, encryption, access control, and contractual bans on sale/monetization/profiling for commercial model training; verify disable rights in incident playbooks.
- Proof of value
- Demo parent decline → alternate assignment without teacher workaround
- Measure time-to-disable under simulated vendor outage
- Audit log of unsupervised access blocks
- Contract exhibit showing no-training clause
Delivery
Role takeaway
- Work
- Rewrite internet-safety policy, build AI registry and consent UX, train staff on risks/limits/compliance, configure kill-switch and substitution pathways, and communicate to families before SY 2027-28.
- Dependencies
- Board policy calendar
- Parent portal vendor
- Curriculum teams for non-AI alternatives
- Legal review of DPAs
- Ownership
- Deputy superintendent / CIO jointly own compliance; principals own building-level unsupervised-use enforcement.
- Skills and adoption
- Mandatory staff PD on AI risks, limitations, safety, and compliance before student-facing enablement; refresh when tools change.
- Governance checkpoints
- Board adoption of amended internet-safety policy
- PreK-5 tool approval log
- Quarterly AI inventory attestation
- Kill-switch tabletop exercise
- Acceptance criteria
- 100% of student AI tools listed with parent-facing disclosures
- Documented comparable non-AI alternative per consented tool
- Admin can disable each tool within agreed minutes without vendor
- Training completion recorded for instructional staff
- Risks
- Official rule text not retrieved (404) may differ in edge cases
- Comparable-alternative requirement may be operationally hard for personalized tutors
- Partial roster consent creates classroom management complexity
Implementation considerations
Lighthouse Advisory interpretation across the operating dimensions a public-sector buyer must settle before this evidence becomes a design. Each note answers the question under its heading for this specific source.
Architecture and integration
What must connect, and where does the AI sit in the workflow?
District stacks need an authoritative AI-tool inventory tied to SIS/parent portals for consent and course-level disclosure; LMS and filtering controls must allow per-student alternative pathways and immediate feature disable independent of vendor consoles; consumer GenAI endpoints should stay blocked for unsupervised student use.
Governance
Who approves, reviews and stays accountable for outcomes?
Fold AI into board-adopted internet safety policy with public review cycles; define who approves PreK-5 tools; document teacher permissions for graded-work exceptions; align academic-integrity defaults (AI off unless instructor enables).
Security and privacy
What data, permissions and controls need testing?
Contract for no sale/monetization/profiling to train commercial models, U.S. data storage preference, incident and disable rights, and COPPA/FERPA-aligned DPAs; companion/SEL chatbot categories become disallowed product classes.
Accessibility and workforce
Who is affected, and what skills or accommodations follow?
Staff must be trained before student use; declined students need comparable non-AI instruction, which pressures accessible alternate materials and teacher workload planning.
Procurement
What should contracts, pricing and exit terms secure?
RFPs should require consent APIs, substitution plans, disable rights, U.S. data preference, and bans on commercial training/monetization; companion/SEL chatbot SKUs are non-compliant product classes.
Operating model
Which teams own the service once it runs?
AI governance moves into the same operating rhythm as internet safety—board cycles, inventories, and building-level enforcement—rather than isolated innovation pilots.
Publication history
- 2026-09-17K–12 · Issue 124 resources
Stable resource ID: florida-boe-ai-parent-consent-rules-20260916