Ten Ethical Principles for AI in Education

Ten Ethical Principles for AI in Education

– Technology that does not run ahead of the child, and does not replace the teacher

1. AI is a supporting tool, not the center of education

AI cannot be the agent of education.

At the center of education there must always be people — above all, the relationship between the child and the teacher.

2. AI teachers are permitted only as a transitional alternative

AI teachers should be introduced only temporarily and in limited ways, to make up for teacher shortages, emotional gaps, and blind spots in learning.

They must be designed as a transitional structure, not a replacement structure.

3. AI does not intervene before the child's language has formed

Before children learn to speak their own thoughts and express them in writing,

exposure to LLMs and automated teacher systems must be limited.

4. AI cannot replace the senses

An education of listening, speaking, writing, making, and waiting

can be passed on only through the flow of sensation that one person conveys to another.

AI cannot be a substitute for sensory education.

5. AI in education must wait for the pace of the child's thinking

Instead of giving fast answers,

AI should become a questioner that draws out the child's thoughts, a mirror, and a helper that waits.

6. AI must be subordinate to the teacher's judgment and emotion

AI must be designed so that it does not make judgments independently,

and should operate only within the teacher's educational philosophy, emotional flow, and relational context.

7. AI does not store the child's data permanently

AI must not permanently collect or analyze a child's learning content and emotional patterns,

beyond de-identified, temporary storage.

An AI that remembers is dangerous.

8. AI cannot be a proxy for emotion in education

Comfort, empathy, and companionship can be imitated,

but only a human being can give true emotional care.

AI is an imitator of emotion, not a subject of emotion.

9. Education policy must consider the child's rhythm before the technology

The timing, the method, and the proportion of adoption

must all be decided on the basis of the child's developmental stage and emotional stability, not on technological progress.

10. Technology does not replace education. It walks alongside it

AI never completes education.

Education is a process of resonance that unfolds in the touch of a hand, the look in the eyes, the tone of voice, and the act of waiting.

Technology may stand beside it,

but it must not become the center.

Ma-eum Company's proposed official position on the teacher system

Public-school teachers today

have no training to teach the senses,

no time to wait for questions,

and no room to watch over a single child for long — that is the structure they live in.

We are not calling into question the ability of teachers,

but the structure of a society that has failed to raise teachers into true teachers.

AI cannot replace teachers,

but the current teacher system, too,

is not ready to teach people in a way that makes them fully human.

Therefore, before the question of whether to adopt AI teachers,

we believe we must first redefine, retrain, and restructure the teacher.

Title:

"A Declaration on the Redefinition, Retraining, and Restructuring of the Teacher"

1. Now, we must redefine the teacher

Today's education

makes us evaluate children too early,

and keeps teachers bound to standards for too long.

I say:

it is now time to redefine the teacher.

A teacher is no longer a manager of knowledge or a controller of exams.

2. A teacher must be a companion in resonance

A teacher is someone who

restores the senses,

protects thought,

and waits for the child's 'slow growth.'

Someone who does not speak the right answer first,

and who is sometimes simply present, without a word.

A teacher does not interfere with the child's 'becoming themselves,'

but stays beside them and flows along with them.

3. I do not trust the current teacher system

Today's public-education teacher system

has neither the room to love a child,

nor the time to watch a single child for long.

This is not the problem of individual teachers.

It is the problem of a system that has failed to raise teachers into true teachers.

I do not believe that teachers today

are ready to teach real human beings.

But I believe that they can be remade.

4. The Ma-eum teacher retraining system – a three-stage proposal

Stage 1. Sensory recovery training

• Nonverbal observation

• The rhythm of waiting

• Resonance training of hand, eye, and ear

A teacher who reads emotion before technology

Stage 2. Respect for the child's language

• Accepting clumsy expression

• Avoiding leading questions

• Saying 'I heard you' rather than 'You're wrong'

Bringing the child's language to life rather than the teacher's

Stage 3. Teacher ethics training for the age of the LLM

• Curbing dependence on AI feedback

• Guarding against over-trust in data

• Training to maintain emotional connection

Forming a healthy distance between people and technology

5. The teacher must change before the technology

Before AI teachers,

we must first create real teachers.

Unless the current teacher system is restored,

no education reform will last.

I believe this.

A teacher who teaches children slowly, deeply, and differently once more.

Only when that teacher begins

does Ma-eum education begin too.

Written by

Lee DongHun (이동훈)

Ma-eum Company – Designer of resonance-based AI systems