Where do your kid’s photos actually go?

Your kid scores on a Saturday morning, you take the perfect photo, and you want to share it. That instinct is good. But it’s worth knowing, concretely, what happens to that photo depending on where you put it — because the answers are different, and most of us never read them.

On the big social platforms

Meta — Facebook and Instagram — states in its own AI documentation that it trains generative AI models on the public posts and photos people share on its platforms (see “Generative AI at Meta”). Content from private accounts and private messages is excluded, per Meta’s statements. That distinction matters — and it cuts both ways. If your profile is public, or you post to a public team page or a public hashtag, photos of your kid can become part of the material an AI model learns from. There is no meaningful way to pull them back out later.

Even fully private sharing on a social platform still lives inside an advertising business. The photo itself may stay between you and your followers, but the activity around it — what you post, when, who engages — feeds systems whose job is to hold attention and sell ads. None of this is a scandal; it is simply what those products are. The question is whether it’s what you want for a childhood’s worth of photos.

What “private by default” looks like instead

playback was built as the opposite kind of product, and the differences are structural, not marketing copy:

Photos of your kid never touch an AI model. Not for training, not for analysis. They live on your device, and — only if you sign in and sync — in a private storage bucket that serves them back to you alone, through expiring signed links.
No feed, no ads, no tracking. There is no public surface at all. Nobody discovers your kid on playback. The only people who ever see a game are the ones you hand a link to, and every link is revocable.
Where we do use AI, we say so plainly. If you tap Polish or use voice capture, the text of your note (never your photos) is processed through OpenAI’s API to write it up. API content is not used to train OpenAI’s models (their data-usage policy), and we don’t store the audio. You can skip these features entirely and the app works fine. The full details are in our privacy policy.

A quick checklist, whatever app you use

1. Check whether your profile — and any team page you post to — is public. Public means available for AI training on major platforms.
2. Ask where photos physically live: on your device, in a private store, or on a platform’s servers under a license you granted by signing up.
3. Look for the business model. If the product is free and runs ads, attention is the product, and the incentives follow.
4. Prefer places where sharing is a deliberate act — a link you send and can revoke — over places where sharing is the default.

The moments are worth keeping. Keeping them shouldn’t mean donating them.