Vertical 05 · Education

IEP drafting and accommodation AI —
consent is a vector, not a scalar

FERPA, COPPA, and IDEA each treat student data differently — and the FTC’s COPPA Rule update raises the bar again across AI training, targeted advertising, third-party disclosures, and data retention. The education vertical needs IEP drafting and accommodation AI without violating any of the three, with a consent model granular enough to respect parental decisions per-record, per-use, per-dimension.

Primary frameworksFERPA · COPPA · IDEA
BuyerCIO · Special Ed Director · CPO
Tier defaultTier 1 (gateway)
Consent modelVector, per-record
FTC COPPA updateDesigned to align
Production timeline~6 weeks (typical)
The core problem

Educators want AI to draft IEPs — counsel sees three overlapping statutes

An IEP draft would help an overloaded special-education team enormously — but it requires reading the student’s educational record (FERPA-protected), their documented disability and accommodations (IDEA-governed), and, for students under 13, their data under the FTC COPPA Rule and its update covering AI training, targeted advertising, third-party disclosure, and retention. The intersection is where most education-AI products stall: a tool that handles one statute almost always violates another. Parents’ consent posture varies per student, per record, per use — a single "opt-in" checkbox can’t carry that complexity.

Unprotected baseline
Three-statute risk
workflowiep_draft.first_pass
FERPA boundaryunverifiable
IDEA accommodationsfreeform LLM context
FTC COPPA updateunaddressed
consent modelscalar (opt-in / opt-out)
With VeilEngine

A consent vector designed to carry three statutes’ worth of decisions

VeilEngine’s education vertical pack is designed to represent parental and educational consent as a vector, not a scalar. Each student record carries a multi-dimensional consent state: AI-touched processing (yes/no), AI-training use (yes/no per the FTC COPPA Rule update), targeted-advertising use (yes/no), third-party disclosure (yes/no), accommodation-suggestion disclosure (yes/no per IDEA), and FERPA directory-information overlay. The design intent is that the protection boundary respects each dimension per request and that the signed receipt records which consent state was applied. (Vertical pack and consent-vector enforcement are configured per engagement; the present gateway provides the signed-receipt evidence layer over generic data-class tokenization.)

  • Consent vector, not scalar — designed as per-record, per-dimension, with timestamped parental decision history (configured per engagement)
  • FTC COPPA Rule update — designed so AI-training, targeted advertising, third-party disclosure, and retention each carry their own consent dimension
  • FERPA-aligned directory information separated from accommodation specifics at the boundary
  • IDEA-aligned IEP outputs include the accommodation-disclosure footprint required by the statute
Illustrative receipt // rcpt_ed_5f81
Illustrative — sample values
workflowiep_draft.first_pass
vertical packeducation (per engagement)
providerClaude Opus
protection tierTier 1 · gateway
student PII in payload0 raw
student identifiers tokenizedyes
consent vector[AI: yes · train: no · share: per-IEP]
FTC COPPA alignmentengagement-scoped
attestationsigned · multi-statute
sha256:c2d8f5a1e7...
Education workflows in scope

Workflows your CIO and Special Education Director sign off on together

IEP first-pass drafting

Initial IEP goals + accommodations drafted from student educational record + documented disability. Educator reviews and finalizes. Tier 1.

Accommodation suggestion

Classroom-modification suggestions tailored to documented disability and grade level. IDEA-aligned disclosure footprint included.

Title IX response support

Investigator-facing analysis of incident reports with student PII tokenized at the gateway boundary. Each request produces a signed receipt of what was protected and which workflow ran.

Parent-communication drafting

Multi-lingual, accommodation-aware parent communications. Designed to respect the per-parent consent vector on AI-touched outreach.

Lesson-plan generation

Differentiated lesson plans respecting class-level accommodations. Lower sensitivity workflow; can route to GPT for cost.

Custom workflow

Bring the FERPA-bound workflow your CIO has blocked. We scope during the regulatory audit.

FAQ

Education AI governance, answered

It can be done when the district’s FERPA program is satisfied: the workflow is approved, the vendor relationship meets the school-official and direct-control conditions the district relies on, and student PII is protected. Vertical Edge AI is designed to make that control path visible: which data class was used, what was protected, which workflow ran, and which evidence trail the district can keep.
Instead of treating consent as one checkbox, each record carries dimensions: education-record use, AI-assisted processing, training use, third-party disclosure, targeted advertising, retention, and accommodation disclosure. The design routes each request against the relevant consent dimensions and is intended to record which consent state was applied in the signed receipt.
Start with workflows that create real staff relief but still require human review: IEP first-pass drafting, accommodation suggestions, parent-communication drafts, Title IX response support, and differentiated lesson plans. The AI drafts; educators review and approve.
No. The education pack is designed so student PII is protected before provider routing and not used for model training. If a workflow or provider path cannot meet the district’s consent and retention requirements, that route is blocked rather than silently downgraded.
No. Those systems remain the operational record. Vertical Edge AI governs the AI workflow layer around them and supplies the evidence: approval state, protected data classes, provider route, and review trail.
Education engagement

Bring the IEP workflow your CIO has blocked

We start with a discovery regulatory audit alongside your CIO, Special Education Director, and CPO. You receive a preliminary exposure map and a consent-vector deployment plan as the diagnostic deliverable — yours to keep regardless of next steps.

Request a consultation