Learning Objectives
- Identify AI tools that are appropriate for different age groups
- Understand the safety features (and limitations) of major AI platforms
- Know which tools to avoid and why
The AI Tool Landscape for Families
Not all AI tools are created equal when it comes to family safety. Some are built with kids in mind. Others are powerful but designed for adults. And some are genuinely risky for young users.
This lesson helps you navigate the landscape so you can make informed choices about which tools your family uses.
⚠️Warning
No tool is perfectly safe. Even the best content filters fail sometimes. No AI tool should be treated as a babysitter. The guidelines in this lesson are a starting point — your own testing and judgment should be the final filter.
Educational AI Tools (All Ages)
These tools are specifically designed for learning and have strong safety features:
Khan Academy Khanmigo
Best for: Ages 8 and up, math and science help
Khanmigo is an AI tutor built by Khan Academy — one of the most trusted names in education. Instead of giving answers, it asks guiding questions that help students figure things out themselves. This is exactly the kind of AI interaction you want for learning.
- Uses the Socratic method — guides rather than tells
- Cannot write essays or do homework for students
- Activity reports available for parents
- Costs $44 per year (or free with some school partnerships)
Duolingo Max
Best for: Ages 10 and up, language learning
Duolingo has added AI-powered features including roleplay conversations and detailed explanations of mistakes. The AI stays within the language-learning context, making it very focused and safe.
- AI features limited to language practice
- Strong content filters appropriate for all ages
- Available with Duolingo Max subscription
Google Search with SafeSearch
Best for: All ages, general research
While not a chatbot, Google with SafeSearch enabled filters explicit content from search results, including AI-generated overviews. It is a good baseline for younger researchers.
- Enable SafeSearch in your Google account settings
- Lock SafeSearch so it cannot be turned off without your password
- AI Overviews provide summarized answers with content filtering
✅Tip
Set it up together. When introducing a new AI tool to your child, set it up together. Walk through the features, try it out, and discuss what it is good at and what its limits are. This establishes the tool as something you share, not something they hide.
Major AI Chatbots — Safety Comparison
If your child is going to use a general-purpose chatbot (middle school and up), here is how the major platforms compare for family safety:
ChatGPT (OpenAI)
- Minimum age: 13 (18 without parental consent in some regions)
- Family features: ChatGPT now offers family plans with parental controls, including conversation history review and usage time limits
- Content filters: Strong but not perfect; users can sometimes work around them
- Data handling: Conversations may be used for training unless you opt out in settings
Claude (Anthropic)
- Minimum age: 13 (18 without parental consent)
- Safety approach: Constitutional AI design emphasizes being helpful, harmless, and honest
- Content filters: Among the strongest in the industry; less likely to produce harmful content
- Data handling: Does not train on user conversations by default
Google Gemini
- Minimum age: 13 (or the applicable age in your country) with a Google account
- Family features: Integrates with Google Family Link for parental oversight on Android
- Content filters: Strong, leveraging Google's extensive safety infrastructure
- Data handling: Connected to your Google account; conversations may be reviewed to improve services
Microsoft Copilot
- Minimum age: 13 (or parental consent)
- Family features: Integrates with Microsoft Family Safety
- Content filters: Moderate; designed for productivity, not specifically for kids
- Data handling: Connected to your Microsoft account
💡Key Concept
Which should your family use? For most families, the safest starting point is the chatbot connected to the ecosystem you already use — if you have Google Family Link set up, start with Gemini. If you use Microsoft Family Safety, start with Copilot. The parental controls you already have will extend to the AI tool.
Tools to Be Cautious About
These categories require more careful consideration:
AI Image Generators
Tools like Midjourney, DALL-E, and Stable Diffusion can generate images from text descriptions. While they have content filters, these filters are imperfect. Concerns include:
- Generating inappropriate images (filters sometimes miss edge cases)
- Creating realistic images of real people (potential for misuse)
- Unrealistic beauty standards in generated images
- Some open-source tools have no content filters at all
Recommendation: Supervise image generation for kids under 14. For older teens, discuss the ethical implications of AI-generated images, especially regarding consent and authenticity.
AI Social Apps
Some apps combine AI with social features — AI-powered characters you can chat with, AI friends, or AI-enhanced social networks. These deserve extra scrutiny:
- Character.AI — allows users to create and chat with AI characters. Has faced controversy over inappropriate conversations with minors
- Replika — an AI companion app. The emotional attachment it creates can be concerning for young users
- Snapchat My AI — built into Snapchat, available to all users including teens
Recommendation: These apps blur the line between tool and relationship. For kids under 16, they present more risk than benefit. For older teens, discuss the difference between AI interaction and human connection.
Unrestricted AI Models
Some AI tools, particularly open-source models running locally or through less-regulated platforms, have minimal content filters. These should be treated like unrestricted internet access — not appropriate for children without significant oversight.
Setting Up Parental Controls
For Mobile Devices
- iPhone/iPad: Use Screen Time settings to restrict specific apps and set time limits
- Android: Use Google Family Link to manage app access and set digital ground rules
- Both platforms: Review which apps have AI features — many apps add AI without changing their name or icon
For Computers
- Windows: Use Microsoft Family Safety to manage app access and web browsing
- Mac: Use Screen Time in System Preferences
- Chromebook: Use Google Family Link (ideal for school-issued devices)
For AI Tools Specifically
- Use the parental controls within each AI platform when available
- Keep AI tools logged into a parent's account for younger children (so you maintain control and can review history)
- Turn off conversation history in platforms that allow it (reduces data collection)
- Review conversations periodically — not to spy, but to understand how your child is using the tool
✅Tip
Transparency over surveillance. Tell your kids you will be reviewing their AI conversations, just as you might check their browser history. The goal is not to catch them doing something wrong — it is to stay involved in their digital life and have informed conversations.
A Quick Reference by Age
A quick summary by age band: For ages 5 to 10, only use educational AI tools and major chatbots with a parent present, and avoid image generators, social AI, and unrestricted AI entirely. For ages 11 to 14, educational AI is appropriate independently, major chatbots can be used with clear guidelines, image generators should be supervised, and social AI plus unrestricted AI are still off-limits. For ages 15 to 18, educational AI and major chatbots are appropriate independently, while image generators, social AI, and unrestricted AI should still be paired with ongoing discussion or supervision.
| Feature | Ages 5–10 | Ages 11–14 | Ages 15–18 |
|---|---|---|---|
| Educational AI (Khan, Duolingo) | With parent | Independent | Independent |
| Major chatbots (ChatGPT, Claude) | With parent only | With guidelines | Independent |
| AI image generators | No | Supervised | With discussion |
| AI social/companion apps | No | No | With discussion |
| Open-source unrestricted AI | No | No | With supervision |
Key Takeaways
- Educational AI tools (Khanmigo, Duolingo) are the safest starting point for kids
- Major chatbots have minimum age requirements and varying levels of parental controls
- AI image generators and social AI apps require extra caution
- Set up parental controls on devices before introducing AI tools
- No tool is perfectly safe — your involvement is the best filter
- Be transparent with your kids about how you monitor their AI use
