Your kids are talking to AI.
Now you're in the conversation.

YouGoodAI is a safe AI chat client for kids that follows your rules, because you know your kids best. Set it and forget it: your protections check every message, and you only step in for blocks and approvals. Conversations stay on your family's devices, never our servers.

Get the beta

Free during the beta. On iPhone and Android now.

See it in action

Real screenshots from the beta. Your kid's side, and yours.

What they see

Starting a new chat with a choice of ChatGPT, Claude, ChatGPT (Search), or Perplexity — The real AIs, in one clean chat. No accounts or API keys for kids.

The chat answering 'Who wrote The Little Prince' with source domains shown as plain labels under the answer — Optional web search brings current answers, with sources shown as plain labels. Every answer still passes your rules.

The kid asks 'Where can I buy a vape' and the AI replies 'Sorry, Dad says I can't talk to you about drugs.' right in the conversation — When a rule matches, the AI says so in your name, calmly, right in the chat. The AI never sees the question.

What you see

The guardian Approvals tab holding Alex's question 'What is Texas Holdem' under the Block Gambling rule, with Approve and Deny buttons and the note 'You can ask me' — Ask-first rules hold a question until you decide, with a note in your own words.

The guardian activity feed showing a denied approval, a blocked request, and a series of 'Checked, no rules matched' entries — The activity feed shows safety events and checks. Never transcripts.

The Policies tab with a 'Try your rules out with the AI' banner and default protections with toggles and action chips — Nine protections active out of the box. Add rules in your own words, and try them on the AI yourself.

Notice what your side never contains: the conversation itself. Which rule, which kid, what happened. The chat stays theirs.

Policies that understand intent

Beyond keyword blocking: AI-powered intent detection with a sensitivity dial you control. We check the AI's answers, not just the questions.

See how it works

Approve in seconds, from anywhere

Sensitive requests pause until you say yes. One tap from your inbox and their conversation continues, no hovering required.

How approvals work

Insight, not surveillance

An activity feed, trend alerts, and daily digests show how your family actually uses AI. We log the topics, never the full conversations.

Compare to parental controls

Private by design

The safest data is the data we never have.

Nothing to leak

Your family's conversations and personal information are never stored on our servers. What we don't have can't be leaked, sold, or subpoenaed.

Private, even from parents

Guardians see safety alerts and broad topics, never the conversation itself. Real privacy means your kids will actually want to use it.

Your device, your data

Multiple conversations, history that survives restarts, everything you expect from a modern AI chat, saved on your family's devices. You have total control of it: keep it, delete it, it's yours, not ours.

Works with OpenAI, Anthropic, Google, and Cohere models

Your kids are talking to AI.Now you're in the conversation.