For technology staff

AI training that works inside your school’s rules.

Child protection comes first. We teach on the tools you’ve approved, and give your tech coaches a way to carry the work into classrooms.

  • WhoTech coaches and IT directors
  • ToolsYour approved systems
  • SetupsLocked down is fine
  • WhereOn your campus
What technology teams face

Four risks, four safeguards.

The concerns IT directors and tech coaches raise first, and how every session is set up around them.

  1. The risk

    Tools arrive faster than you can review them

    Teachers find new AI tools on their own, and each one is another review.

    How we handle it

    No new tools introduced

    We review your approved list with you before planning. Sessions run on the systems your school already allows.

  2. The risk

    Student data and child protection

    Every new tool raises the same questions about what data goes in and who sees it.

    How we handle it

    No student records in sessions

    The data rules are agreed with your team first, and teacher sessions use teachers’ own materials, not student records.

  3. The risk

    Uneven adoption across staff

    A few teachers move fast while most wait for something they can use in class.

    How we handle it

    Your coaches in the room

    Sessions run on your campus with your tech coaches there, so the people who support teachers see the work first-hand.

  4. The risk

    Every tool change restarts training

    When a product changes or is replaced, the training built around it goes with it.

    How we handle it

    What stays when tools change

    Your coaches keep the session plans, teacher examples, and shared templates, and the method carries to the next tool.

Student data comes first

Before which tool, ask: whose account is it?

Our first job is safeguarding students. The same AI tool can be safe or unsafe for student data, depending on the account it runs in.

A personal account

Signed up by one person

Terms
Consumer terms, accepted by one user
Data agreement
None with the school
Safeguards set by
The provider
Student data
Not approved for student data

A school-approved platform

Set up and managed by the school

Terms
An enterprise agreement with the school
Data agreement
Yes, agreed by the school
Safeguards set by
The school
Student data
Approved, under the school’s policy
How we handle it
  • Inside your approved platforms

    We work only in the accounts and tools your school approves, and agree the data rules with your team before any session.

  • No student data in personal accounts

    Student names, work, and records never go into a personal or consumer account.

  • Sample data for practice

    Teacher sessions use teachers’ own materials or anonymized sample data, not student records.

  • AI use disclosed

    Learners say which steps they handed to AI, and how they used it.

  • Teachers learn to check the terms

    Every teacher session covers whose account a tool runs in and what its data agreement says.

  • Your data-protection lead decides

    We follow your data-protection officer and policy, and adjust when they do.

The law behind it
  • China

    Personal Information Protection Law (PIPL), in force since Nov 1, 2021

    Personal information of children under 14 is sensitive personal information (Art. 28). Processing it needs the consent of a parent or guardian, and dedicated processing rules (Art. 31).

    Official text (Chinese) →
  • European Union

    General Data Protection Regulation (GDPR)

    Depending on its location and families, a school may also fall under the GDPR. For online services offered directly to a child, consent below 16 must come from a parent; member states may set 13 to 16 (Art. 8).

    Official text (EUR-Lex) →
  • Thailand

    Personal Data Protection Act B.E. 2562 (2019)

    The data controller is whoever has the power to decide how personal data is collected, used, or disclosed (Section 6). For student data, that is the school.

    Personal Data Protection Committee →

General information, not legal advice. Your school’s data-protection lead decides what applies.

Before adopting any AI tool

Eight questions to ask first.

  1. Whose account is it?A personal account, or one the school approves and manages?
  2. Is there a data agreement with the school?Signed by the school, not accepted by one user.
  3. Where is data stored and processed?And does that fit the laws your school follows?
  4. Will minors use it?If so, is parental consent in place where the law requires it?
  5. How long is data kept?Can the school delete it, and how?
  6. Does the provider train on what users type or upload?Can that be switched off for the school?
  7. What can school administrators control?Accounts, sharing, and activity logs.
  8. How will families be told?Which tools are used, and for what.
What you get

Built for the setup you already run.

We value child protection and student data privacy, and we can build for your school’s locked-down requirements.

What the research says

Use is rising fast. Policy and training lag.

Four findings from RAND’s national surveys of districts, principals, teachers, and students. Each links to its source.

How we plan with you

Your rules first, then the sessions.

We agree on tools and data with your team before any teacher or student session.

  1. 01 · Review

    Your tools and rules

    We go through your approved systems and child-protection policy with you.

  2. 02 · Set up

    Inside your stack

    Every session and example is built on those systems.

  3. 03 · Run

    With your coaches

    Sessions run on your campus, with your tech coaches in the room.

  4. 04 · Hand off

    Ready to run again

    Your coaches keep the session plans and teacher examples.

Where we have done this

Run inside real institutions.

  • Tsinghua University · Staff training

    Schwarzman College

    Training for staff built around their own work, with their judgment in charge. The same format as AI for Educators.

    Who
    Staff, Schwarzman College
    Format
    Hands-on workshop
    Taught by
    Phillip An
    See results →
  • International School of Beijing · On campus

    A five-day student program

    Over five days, students built working websites and apps with AI and presented them to judges, on the school’s campus with mentors in the room.

    Where
    International School of Beijing
    When
    Summer 2026
    See results →
  • Next · ACAMIS Technology Conference

    Keystone Academy, Beijing

    A 60-minute hands-on workshop for teachers and tech coaches, built on the case study from the International School of Beijing.

    When
    Sunday, October 18, 2026
    For
    Teachers and tech coaches
    Session details →
Tools we have worked with, when a school approves them
  • ChatGPT
  • Claude
  • DeepSeek
  • Kimi

Plus Copilot, Doubao, and others your school allows. We review them with you before planning any session.

A member of
  • AnthropicStartup Program
  • OpenAIPartner Network
Our team studied at
  • International School of Beijing
  • Brown University
  • Caltech
  • Harvard University
  • Tsinghua University
Worked at
  • McKinsey & Company
  • Goldman Sachs
  • Y Combinator
  • Apple
  • Microsoft AI
  • Deloitte
  • Kimi (Moonshot AI)
Fair questions

What technology teams ask.

We can’t use frontier models. Does that rule this out?

No. We build for your school’s requirements and approved systems, such as a locked-down Copilot. What teachers learn is a method, so it works on the tools you allow.

How do you handle student data?

We agree the data rules with your team before any session. Teacher sessions use teachers’ own materials, not student records.

Which tools do you use?

Whichever your school approves: Copilot, tools that work in China such as Kimi, Doubao, and DeepSeek, or others. We review them with you before planning any session.

Who supports it after you leave?

Your tech coaches, with what stays behind: session plans, teacher examples, and shared templates. Because teachers learn a method rather than one product, there is less to troubleshoot when tools change.

Which privacy laws apply to us?

It depends on where your school is and who your families are: for example PIPL in China, the GDPR for EU families, or PDPA in Thailand. Your data-protection lead decides; we work inside their policy.

Do teachers need new accounts?

No new tools are introduced. Sessions run on the systems your school has already approved.

Next step

Check fit with your setup.

Tell us which tools your school approves and your rules on student data. We reply within three working days.