AI-Resistant Education for Kids – Concept Document

1. Motivation and Mission

Core motivations

Protect and strengthen children’s human capacities (curiosity, empathy, agency, critical thinking) in an AI-saturated world.
Intervene before harm or misuse, similar to good sex education or digital safety education: assume contact with AI is inevitable, and focus on prevention and preparation, not fear or bans.
Offer families and schools a sane middle path between “ban all AI” and “let kids figure it out alone.”
Teach kids to see AI as a powerful tool/“creature” they must learn to handle safely, not as a friend, a god, or a replacement for their own thinking.

Mission Statement

We help children become confident, resilient AI users—developing judgment, integrity, and human skills that no technology can replace.

We teach kids and adults to see AI as a powerful tool to be used wisely, not a replacement for thinking, relationships, or responsibility.

2. Core Organizational Values

2.1 Child-centered dignity

Children’s safety, privacy, emotional well-being, and long-term development come before convenience, profit, or hype.
The organization treats every program as serving children’s interests first, not vendors or systems.

2.2 Human agency over automation

AI should augment, not replace, human judgment, relationships, and responsibility.
Kids learn that humans remain in charge; they are training the AI “dragon,” not obeying it.
The org pushes back against uses of AI that remove meaningful human oversight.

2.3 Honesty and transparency

Be explicit about:
- When AI is used.
- What AI can and cannot do.
- Where its limits and risks are.
Model and teach kids to be transparent: “I used AI to help with X,” instead of hiding or cheating.

2.4 Equity and inclusion

Design materials that work across cultures and income levels.
Avoid deepening digital divides or reinforcing bias.
Teach children about fairness and how AI systems can inherit or amplify human bias.

2.5 Privacy and minimal data

Collect and store as little child data as possible.
Avoid sending identifiable student data to external AI services when possible.
Prefer tools and practices that respect children’s privacy and do not profile or manipulate them.

2.6 Critical compassion

Combine critical thinking (question, verify, resist manipulation) with compassion (understand effects of AI on people, especially vulnerable groups).
Encourage kids to think about how AI impacts themselves, their friends, and society.

2.7 Evidence-informed practice

Base curricula and policies on:
- Child development research.
- AI ethics and responsible AI frameworks.
- Real classroom experience and feedback.
Commit to revisiting and updating programs as evidence and technology change.

3. Guiding Principles for Programs and Partnerships

3.1 Human-first, AI-aware design

Every program has explicit developmental goals (empathy, critical thinking, self-regulation). AI is context, not purpose.

3.2 Age-appropriate, staged learning

Content evolves from concrete (grades 3–5) to complex (grades 6–8+). Themes repeat with increasing nuance: privacy, honesty, boundaries, identity.

3.3 Prevention over punishment

We prepare kids before problems arise. When misuse happens, we focus on reflection and learning, not shame.

3.4 True partnership with adults

We provide clear guidance, sample policies, ready-to-use lessons, and scripts. We help educators talk about AI as they would sex ed: early, honest, calm, value-based.

3.5 Vendor independence

We maintain independence from specific AI companies. We disclose all partnerships and avoid endorsing tools for money or influence.

3.6 Walk the talk

Our staff use AI transparently. We include youth voices in design. We review our own practices regularly to stay true to our values.

4. Conceptual Frame: AI as “Another Entity”

4.1 AI as a new kind of “other”

Kids already learn different interaction rules for:
- People (mutual respect, feelings, consent).
- Animals (body language, care, boundaries).
AI is “another entity”:
- Human-like in language and style.
- Not human in feelings, responsibility, or consciousness.
- Requires its own interaction rules.

4.2 Desired social skills with AI

Kids should learn to:

Recognize what AI is and is not:
- “It can sound smart, but it doesn’t actually know or care like a person.”
Follow “conversation rules” with AI:
- What’s safe to ask.
- What is unsafe (private info, intimate topics, self-harm, hatred).
- How to stop and ask a human if something feels wrong.
Manage emotions and attachment:
- Name the feeling that AI seems like a friend.
- Understand that the system does not have genuine care or responsibility.
- Prefer human relationships for emotional and moral questions.

5. Metaphor: “Train Your AI Dragon”

5.1 Why this metaphor

AI is like a dragon:
- Powerful, surprising, helpful, and sometimes dangerous.
- Follows training and patterns, not morals or feelings.
Kids are the rider/trainer:
- Learn commands, boundaries, and safety rules.
- Know when NOT to bring the dragon into a situation (e.g., private, emotional, or integrity-sensitive tasks).
Fits well with:
- Children’s imagination.
- The idea of discipline, responsibility, and growth.
- The sex-ed analogy: you can’t pretend dragons don’t exist; you prepare kids to handle them.

5.2 Example lesson concepts (high-level)

Not full curriculum, just the framing:

Lesson: “Meet your dragon”
- What can it do?
- What can’t it do?
- How is it different from a friend or a pet?
Lesson: “Safety rules for dragon riders”
- 4–5 simple rules: no private info, double-check facts, ask adults, don’t use it to hurt others, don’t let it do your whole work.
Lesson: “Dragons don’t have hearts”
- AI doesn’t feel love, guilt, or responsibility.
- Humans are responsible for actions and choices.

These sessions would be adapted by age band (3–5, 6–8 grades, etc.).

6. Alignment with Responsible AI Principles

The organization’s values should align with established responsible AI frameworks (e.g., OECD AI Principles, UNICEF AI for Children guidance, TeachAI, etc.).

6.1 Values ↔ Principles mapping (conceptual)

Child-centered dignity
↔ Child rights, safety, and well-being.
Human agency over automation
↔ Human oversight and agency (humans remain in control of important decisions).
Honesty and transparency
↔ Transparency and explainability (clear about AI’s role and limits).
Equity and inclusion
↔ Fairness, non-discrimination, inclusive access; avoid systems that deepen inequality.
Privacy and minimal data
↔ Data protection and privacy; minimize data collection and risk.
Evidence-informed practice
↔ Accountability and continuous improvement; review impact, adjust with new evidence.

6.2 Translating alignment into policies

Examples:

Tool vetting:
- No use of tools that profile or manipulate children, or that monetize child data in exploitative ways.
Human oversight:
- Avoid recommending systems that automate grading or discipline without real human review.
Transparency:
- In all materials and examples, clearly label where AI is used.
Equity:
- Always provide low-tech or no-tech versions of activities where possible.
Privacy:
- Provide simple privacy checklists to schools and parents; encourage minimal data practices.

6.3 Governance loop

Create a small advisory group:
- Educator, child psychologist, AI/ethics expert, youth voice.
Give this group a specific remit:
- Review new programs and partnerships against the values ↔ principles mapping.
Schedule periodic reviews:
- Once or twice a year, check whether programs still align with child-centered, responsible AI standards.

7. Target Age Groups and Interventions

7.1 Elementary school (approx. 8–11)

Focus:

Feelings, safety, simple habits and rules, basic understanding of what AI is.
Very concrete metaphors (dragons, robots, creatures).

Content themes:

“What is AI?”
“When should I ask a person instead of a machine?”
“What is a secret? What is private?”
“How do I know if something is true?”

7.2 Middle school (approx. 11–14)