Know what AI is telling your students before they read it
Guardia™ checks every AI conversation in real time, applies your grade-level rules, and gives parents a clear view of what their child's AI tools say.
The Problem
Students already use AI. Almost no one can see what it says back.
Students use AI to ask questions, get homework help and explore ideas. They often do it with no visibility for parents or schools into the answers. Districts usually find out about a harmful exchange after it happens, if they find out at all. As more states weigh AI child-safety requirements, “we trust the vendor” won’t hold up as an answer.
How Guardia works
A safety check on every message, before the student sees a reply.
- Connect. One integration puts Guardia between students and any AI app, in any language and with any model.
- Scan. Every message is checked against the policy for that student’s grade band: K-5, 6-8 or 9-12.
- Act. Content is blocked, flagged or passed before the student sees the reply, and every scan is logged.
- Inform. Parents and administrators can read the full transcript of any flagged conversation and act on it.
The Platform
Product | For | What it does |
GuardRail (AI safety API) | Districts and ed-tech developers | Scores every AI interaction in real time. District admins can edit K-5, 6-8 and 9-12 rule sets live without a redeploy. Keeps a full audit trail. |
TrustEd (parent dashboard + SDK) | Parents, schools, ed-tech vendors | One dashboard for flagged conversations, one-tap consent approvals, real-time alerts, and limits on time and subjects. The white-label SDK drops into any LMS or app. |
Guardia companion app | Families | Pairs a child’s device with the parent dashboard. On the App Store and Google Play from October 3, 2026. |
Who It's For
- School districts and schools: set age-appropriate AI rules once, enforce them across every AI tool, and answer parents and board members with records instead of reassurance.
- Ed-tech and AI developers: add child-safety scanning and parental consent to your product without building your own safety stack, and give district buyers the answer they’re starting to require.
- Parents: see what your child’s AI tools say, approve consent requests from your phone, and set limits yourself. No technical setup needed.
The AI-Ready Safety Grant
A non-profit grant to put AI safety in place this school year.
CCF is funding a founding cohort of 50 districts and 10 ed-tech partners to deploy Guardia.
| Track A: Districts and schools | Track B: Ed-tech and AI developers |
Who applies | Public districts, charter and private schools, state agencies and regional service centers | Companies whose AI products serve K-12 students |
What’s covered | GuardRail policy engine, TrustEd parent dashboard, companion app for families, policy setup, staff onboarding, audit reporting | GuardRail API and TrustEd SDK integration, integration support, founding-partner status |
Grades | K-8 in Phase I, K-12 in Phase II | K-12 |
Testimonial = Voice From The Pilot
“Guardia makes using AI feel safer. I can ask questions, explore new ideas, and learn with confidence.”
— Second-grade student, Guardia pilot program
For Parents
Want Guardia at your child’s school?
Download the Guardia companion app, or send your principal a note about the AI-Ready Safety Grant.
FAQs
Who can apply?
US public school districts, charter and private schools, state education agencies, regional service centers, and ed-tech companies whose AI products serve K-12 students.
Do we have to switch AI tools?
No. Guardia works with the AI tools you already use, including ChatGPT, Gemini, Claude and tools built into your curriculum.
Is student data sent to AI companies?
Guardia's safety check runs inside Guardia itself. No student message is sent to an AI provider to decide whether it is safe.
Which grades are supported?
Phase I covers K-8. Phase II extends to K-12. Each grade band has its own rules, and your administrators can edit them.
How long does setup take?
For districts, our team handles policy setup and onboarding with you. For developers, GuardRail is a single REST endpoint that works with any language and any model.
Can parents apply?
The grant goes to schools and developers. Parents can download the companion app or use "Tell my school" above.
Why is a non-profit doing this?
Because the organization checking AI for children shouldn't be selling the AI. CCF has spent 28 years in education technology as a 501(c)(3), with no ties to any AI provider.
60 grants. Applications close October 30th, 2026.
