Skip to content

InEd AI Use Guideline

  1. 🔴 Children’s and staff personal data never goes into AI.
  2. 🟠 Remove names before using work data.
  3. ✅ Use InEd-approved tools and accounts.
  4. 🔍 CHECK every output: facts, errors, language, translation, human sign-off.
  5. 🧠 Think first, then ask AI. People make the decisions.
  6. 🗣️ Be open about using AI.
  7. 🚨 Mistake? Tell [the AI focal person] the same day.

In our November 2026 survey of 28 staff, 89% of staff worried that AI gives wrong information, 75% worried about the privacy of office data, and 57% worried about losing their own thinking skills. This guideline answers those three worries with simple rules everyone can follow.

InEd works with migrant children, families and teachers, many in vulnerable situations. Protecting their information comes before saving time.


Rule 1: Know your data (Red, Amber, Green)

Section titled “Rule 1: Know your data (Red, Amber, Green)”
What it includes What you can do
🔴 RED: Never Children’s or families’ names, dates of birth, ID or document numbers, photos or videos, addresses, phone numbers · OOSC case records · health, protection, disability or legal-status information · staff HR files, salaries, performance reviews · passwords, bank details · confidential contracts and anonymous donors’ names Never put it into any AI tool, including paid ones. No exceptions.
🟠 AMBER: Clean it first Survey and Kobo data · attendance and enrolment lists · meeting notes that mention people · draft reports with case examples · partner emails Remove names and identifying details first (replace them with codes such as Child A or School 3), or use only InEd’s approved workspace. If in doubt, treat it as RED.
🟢 GREEN: OK Public information · general writing and emails without personal data · training materials · published reports · your own ideas and drafts Use freely in approved tools.

Quick test: “If this text appeared on Facebook tomorrow, could anyone be harmed or identified?” If yes, it is RED or AMBER.

  • Use the AI tools and accounts InEd approves: Gemini and NotebookLM with your InEd Google account [to be confirmed by Management]. See Which tool should I use?.
  • Personal free accounts are for GREEN data only.
  • Do not install AI browser extensions or meeting recorders that join calls automatically without approval from [Management].

Rule 3: Check before you use it (the CHECK routine)

Section titled “Rule 3: Check before you use it (the CHECK routine)”

AI can sound confident and still be wrong. Before you use any AI output:

  • C: Compare facts. Check every number, name, date and quote against the original source.
  • H: Hunt for errors. Ask the AI: “What might be wrong or missing in your answer?”
  • E: Evaluate the language. Is the tone right for the audience? Is the register formal for government partners?
  • C: Confirm translations. Translate it back into the original language and compare. A native speaker checks anything that goes outside InEd.
  • K: Keep a human sign-off. You are responsible for anything you send, not the AI.
  • Think first, then ask AI. Write your own outline, key points or answer first, then ask AI to improve or challenge it.
  • Use AI as a critical friend: “What is weak in my argument?”, “What have I not considered?”
  • Do not use AI to make decisions about people, such as hiring, admission, or which child gets support. AI can help prepare information, but people decide.
  • Tell your supervisor when a significant piece of work was drafted with AI.
  • Follow each donor’s rules on AI use in proposals and reports. Check before submitting: [funder-specific notes].
  • Do not present AI-generated images of real places, children or events as real photos.
  • Respect copyright: do not upload other people’s paid or confidential materials without permission.
  • Never upload photos or videos of children to AI tools, including for editing or captions.
  • AI-generated images for InEd materials must not look like identifiable real children, and must be labelled “AI-generated illustration” where it matters.
  • Follow InEd’s child safeguarding policy for all visual materials: [link].

If you put RED data into an AI tool by mistake, or AI output caused a problem:

  1. Stop. Delete the chat or file if you can.
  2. Tell [AI focal person / Management] the same day. You will not be punished for reporting honestly.
  3. We will record it, learn from it and improve this guideline.

Questions: ask your team’s AI champion or [the AI focal person].