Privacy & Data Practices

How Prompt-Ed handles your data.

Short version: we don't collect it. Everything you type into a Prompt-Ed tool stays in your browser, on your device, and disappears when you close the page. Here's exactly how that works.

🔒

We don't collect your data. Period.

Prompt-Ed tools run entirely in your browser. When you type into the Student Prompt Builder, the AI Use Documentation Tool, or any other interactive tool on this site:

  • Nothing is sent to our servers. There is no database, no backend, and no API receiving your input.
  • Nothing is saved. When you close or refresh the page, everything you typed is gone.
  • No accounts are required. We don't ask for your name, email, school, or any identifying information.
  • No cookies track your activity. We don't use analytics cookies, session cookies, or tracking pixels.
What this means in practice: If you close the browser tab before copying your work, it's gone. We can't recover it because we never had it.
👁️‍🗨️

No tracking. No analytics on your input.

We don't use Google Analytics, Meta Pixel, Hotjar, or any other service that watches what you type, click, or scroll. There are no third-party scripts on this site that collect behavioral data.

We don't know how many students used the tools today. We don't know what assignments they worked on. And that's by design — we built it this way because student work is not our data to have.

🛡️

Content filtering runs locally in your browser.

Our student tools include a content filter that checks for inappropriate language (profanity, slurs, sexual content, threats, and similar material). Here's how it works:

  • The filter runs in your browser. It's a list of terms checked by JavaScript on your device. Nothing is sent anywhere to be "reviewed" or "flagged."
  • No one is notified. If the filter catches something, you see an error message. That's it. No alert goes to a teacher, administrator, or Prompt-Ed.
  • It's keyword-based, not AI-powered. The filter uses word-boundary matching against a curated list. It's designed to catch clearly inappropriate content while allowing legitimate academic topics (like discussions of historical violence, drug policy, or social justice).
  • It's not perfect. No keyword filter catches everything, and creative misspellings or coded language may get through. The filter is a speed bump, not a wall — it exists to remind students that these are academic tools.
If the filter flags your input and you believe it's a mistake, the error message includes a link to contact us. We'll review the term and adjust if needed.

What happens when a student types something inappropriate.

If a student enters content that triggers the content filter — profanity, slurs, threats, sexual language, or similar material — here is exactly what happens:

  • The tool blocks the input immediately. The student sees an on-screen message telling them the content was flagged and cannot be submitted.
  • No one else is notified. No alert is sent to a teacher, school administrator, parent, or Prompt-Ed. The interaction stays between the student and their screen.
  • Nothing is logged or recorded. Because the tools run entirely in the browser, the flagged input is never sent to a server. There is no record of what was typed — not on our end, and not anywhere else.
  • The student can try again. The filter is a guardrail, not a punishment. Students can edit their input and continue using the tool.
Why it works this way: The filter exists to keep the tools focused on academics — not to monitor or discipline students. A student who tests a boundary gets a clear "no" from the tool and nothing more. This is intentional: these are learning tools, not surveillance tools.
🙋

Why we ask you not to enter personal information.

Our tools remind students not to type in names, email addresses, student IDs, or other personal details. There are two reasons:

  • AI tools save your prompts. When you copy a prompt from Prompt-Ed and paste it into ChatGPT, Gemini, or Claude, that AI platform may store your conversation. Personal details in a prompt become part of a dataset you don't control.
  • Shared documents can be seen by others. When you paste your AI documentation into a Google Doc or similar platform to turn in, anyone with access to that document can see what you wrote. Keeping personal details out protects you.

Prompt-Ed tools don't need your personal information to work. We ask you to leave it out not because we'd misuse it — but because the next tool in the chain might.

📋

FERPA and COPPA alignment.

Prompt-Ed is designed with federal student privacy law in mind:

  • FERPA (Family Educational Rights and Privacy Act): Because Prompt-Ed does not collect, store, or transmit student data, there are no education records to protect, share, or breach. No student data passes through Prompt-Ed's infrastructure.
  • COPPA (Children's Online Privacy Protection Act): Prompt-Ed does not collect personal information from children under 13 — or anyone. No accounts, no forms that collect PII, no persistent data of any kind.

Schools and districts evaluating Prompt-Ed for classroom use can confirm: there is no data processing agreement needed because there is no data processing.

For administrators: Prompt-Ed's tools are static HTML/CSS/JavaScript files hosted on a standard web server. No backend application processes student input. The content filter and interactive builders execute entirely client-side. A technical review of the source code confirms this.
👩‍🏫

Educator tools follow the same rules.

The K-12 Educator Prompt Library, Assignment Builder, and all educator-facing tools on Prompt-Ed operate the same way: entirely in the browser, with no data collection.

Every educator prompt template includes a built-in FERPA/COPPA reminder and constraints instructing AI not to include personally identifiable information. These guardrails are part of the prompt itself — they travel with the teacher into whatever AI tool they use.

🧪

brink. pilot: anonymized session review.

During the brink. pilot period, anonymized session transcripts may be reviewed by the Prompt-Ed team to improve coaching quality. No names, emails, or identifying information are collected or attached to these transcripts.

This review process is used solely to refine brink.'s coaching behavior — how it asks questions, when it scaffolds, and where students get stuck. Once the pilot concludes, this practice will be re-evaluated.

Questions or concerns?

If you have questions about how Prompt-Ed handles data, or if you believe the content filter flagged something inappropriately, please reach out through our contact form.

Last updated: April 2026