AI governance and student data privacy for K–12 schoolsKnow what AI tools are connected before they become a risk.

K12 Guardrail helps school districts identify AI-enabled tools connected to district accounts, understand what permissions those tools hold, and see where student data exposure deserves attention.

  • No software installed
  • No student records needed
  • Built for lean district teams
Sample page from a findings report A findings table listing four AI-enabled apps with what each can access, whether the vendor may train on district data, and a priority. Below it, a cited vendor clause with its document version and review date. Findings report Sample district, for illustration only App Can access Trains on data Priority Lesson planner Full Drive access Yes, opt-out First Quiz generator Class rosters Unclear First Writing helper Drive files No Next Slide designer Creates files No Monitor “...we may use content you provide to develop and improve our models and services...” Lesson planner Terms of Service, v4.2, reviewed Sept. 2026
Illustrative example. App names, findings, and clause are not real.

Four questions every district should be able to answer

  1. What AI tools are connected?Including tools authorized directly by staff or students.
  2. What can they access?Permissions, connected data types, and account reach.
  3. Can vendors train on the data?Based on the vendor's own terms, policies, and agreements.
  4. What deserves attention first?Exposure ranked by sensitivity and governance relevance.

The problem

Districts usually know what they bought. That is not always the same as knowing what is connected.

Teachers and students can authorize third-party applications through district Google or Microsoft accounts. Those connections may never appear on a traditional purchased-software list, even when they hold meaningful permissions.

Shadow adoption

AI-enabled applications can enter the environment outside the normal purchasing process.

Changing vendor terms

Privacy policies, data processing terms, and AI training language change over time.

Limited capacity

Small or shared technology teams may not have time to research every connected application and vendor clause.

The starting point

AI Tool Exposure Review

A plain-language review of which AI-enabled tools are connected to district accounts, what they can reach, what their vendors say about model training, and which governance frameworks may be relevant.

The goal is not to hand the district a generic risk score. It is to create an evidence-backed exposure map that helps district leadership decide what deserves attention first.

What you receive

Inventory of connected AI-enabled applications

A list of the AI-enabled tools that staff or students have authorized through district accounts, including ones that never went through purchasing.

Permissions and account reach for each application

What each tool is allowed to read or change, and how many district accounts have granted it access.

Vendor AI-training language with the specific clause, document version, and review date

The exact vendor language on model training, cited so your team can see what the finding is based on.

Prioritized view of where student data may be exposed

Tools ordered by the sensitivity of the data they can reach, so leadership knows where to look first.

Relevant governance and privacy frameworks to consider

Frameworks that may be implicated by the data a tool can access, such as FERPA, COPPA, and your state's AI requirements, for district counsel or advisors to evaluate.

Questions and follow-up items to take back to vendors

Specific questions to raise with each vendor where terms are unclear or exposure is significant.

Simple process

Designed to minimize the burden on your team.

  1. Export the authorization listAbout 30 minutes

    Your Google Workspace or Microsoft administrator exports the district's third-party application authorizations. We can walk through the process together on a screen share.

  2. Vendor researchTypically 1–2 weeks

    K12 Guardrail reviews the vendor's terms, data processing agreement, privacy policy, and relevant AI language, recording the evidence behind each finding.

  3. Walk through the findingsAbout 60 minutes

    You receive a written report and a review session focused on what is exposed, what to address first, and what questions to ask each vendor.

Data minimization

We do not need student records to do this work.

The review is based on application authorization information: application names, permissions, and the number of accounts using them.

What we need
  • Application names
  • Permissions/scopes
  • Number of authorized accounts
What we do not need
  • Student records or grades
  • IEPs or special education documents
  • Rosters, attendance, or health information
  • Login credentials

Who this is for

Especially useful for districts with small, shared, or stretched technology teams.

K12 Guardrail is being built around the reality that many school districts do not have a dedicated AI governance, privacy, or GRC function. The goal is to give district leaders a clear view of their current exposure without adding another major implementation project.

  • Small districts
  • Shared IT services
  • Google Workspace or Microsoft 365
  • Early AI governance programs
  • Limited internal research capacity

About K12 Guardrail

Built for the gap between classroom innovation and district governance.

K12 Guardrail helps school districts make AI governance practical. The focus is not on slowing down teachers or blocking useful technology. It is on giving district leaders the visibility and evidence they need to understand what is already connected, where student data may be exposed, and what deserves attention first.

The company is being built specifically for K–12 environments, where technology decisions sit at the intersection of instruction, student privacy, vendor management, cybersecurity, policy, and limited staff capacity. The goal is to turn complicated technical and compliance questions into clear, usable information for superintendents, technology leaders, and school administrators.

Practical
Clear findings and next steps, not a generic risk score.
Evidence-backed
Vendor findings tied to the source language, document version, and review date.
K–12 focused
Built around how schools actually adopt, manage, and govern technology.

FAQ

What district leaders usually want to know

Is this a compliance audit?

No. The AI Tool Exposure Review is an exposure map, not an audit, certification, or legal opinion. It identifies connected tools, relevant vendor terms, and areas that may warrant follow-up.

Are you making a legal determination about FERPA, COPPA, IDEA, or state law?

No. The report can identify which frameworks may be implicated by the data a tool can access, but district counsel or other qualified advisors make legal compliance determinations. For background, see our plain-language guide to student data privacy laws.

Does our state require a district AI policy?

A growing number do. Ohio, Tennessee, Arkansas, Maryland, Oklahoma, Utah, Virginia, North Carolina, Idaho, Texas, and Florida now require districts to adopt AI policies through a law or state rule, and most other states, including Michigan, have issued voluntary guidance. See where your state stands in our state-by-state AI guidance tracker.

Do you need access to our Google or Microsoft environment?

No direct system access is required for the review. The district administrator provides the relevant third-party application authorization export.

Why does the report include dates and document versions?

Vendor terms change. Each determination is tied to the source language, document version, and date reviewed so the district can see what evidence supported the finding at that point in time.

How do we send you the authorization export?

On a short video call, we walk your Google Workspace or Microsoft 365 administrator through creating the export. Before anything is sent, we review it together to confirm it contains only application names, permissions, and the number of authorized accounts. Your administrator then uploads it to a restricted folder that only K12 Guardrail can access. We do not accept export files by email.

What if the export includes student or staff information?

It shouldn't, and we check together on the call before it is uploaded. If any column lists individual names or email addresses, we remove it first. If we ever receive more than we need, we delete the extra information right away and let you know.

How long do you keep our files?

Export files are deleted 30 days after we deliver your report, or sooner if you ask, and we confirm the deletion in writing. We keep a copy of the final report for one year so we can answer follow-up questions, then delete it. You can ask us to delete it earlier at any time.

Do you put our data into AI tools?

No. The research and the analysis of your export are done by K12 Guardrail directly. AI writing tools may help draft or format the final report, using settings that keep the provider from training on that content. Your export files are never uploaded to an AI tool.

What happens if the review finds something concerning?

The report is designed to help the district prioritize next steps and identify the questions that should be taken back to the vendor. The purpose is visibility and informed action, not blame.

Pilot program

Help shape K12 Guardrail while getting a clear view of your district's AI exposure.

We are looking for a limited number of K–12 districts to participate in a no-cost pilot review. You receive the findings either way, with no obligation to continue afterward.

Request a pilot review

Prefer a quick introduction first? Email nick@k12guardrail.com.