What Do AI Learning Apps Collect About Your Child? A Parent's Privacy Guide

When your child talks to an AI learning app, they're sharing more than you might think. Their voice. Photos of their homework, sometimes with their name at the top of the page. What they struggle with, how long they take and when they get frustrated. For a general-purpose chatbot, add everything else they choose to type.

Most of this is needed to make the app work. The question is what happens to it afterwards: how long it's kept, who sees it and whether it's used for anything other than helping your child learn.

Why AI apps collect more than older apps

A traditional math app records answers and scores. AI apps often collect much richer data:

• Voice recordings. Voice tutors need to hear your child. The recording may be processed and discarded, or stored.
• Conversation logs. Every question your child asks and every answer the AI gives.
• Images. Photos of homework, which can include names, school names and handwriting.
• Behavioral signals. Response times, hesitation, frustration, topics avoided.
• Model training. Some services use conversations to improve their AI models, which means your child's words may be seen by reviewers or influence future outputs.

None of this is sinister by default. But children can't meaningfully consent to it, which is why the law puts parents in charge.

What the law says

United States: COPPA

The Children's Online Privacy Protection Act applies to online services directed at children under 13, and to general services that know a user is under 13. Its core requirements:

• Verifiable parental consent before collecting personal information from a child.
• Voice counts. Audio files containing a child's voice are personal information under the rule.
• Parental rights to review the information collected, and to have it deleted.
• Data minimization. Services can't require a child to hand over more information than is reasonably needed for the activity.

The FTC finalized significant updates to the COPPA rule in 2025. Among other changes, services now need separate parental consent before disclosing children's data to third parties such as advertisers, and must keep a written data retention policy rather than holding children's data indefinitely. The compliance deadline for most provisions was April 2026.

The FTC takes voice data seriously. In 2023, Amazon agreed to pay $25 million to settle charges that it kept children's Alexa voice recordings for years and undermined parents' deletion requests.

Europe and the UK

Under GDPR, children can consent to online services themselves only from age 16 by default, though countries can lower that to as young as 13. Below that age, parental consent is required. The UK's Age Appropriate Design Code, also called the Children's Code, requires services likely to be used by children to default to high privacy settings and collect only what they need.

Teenagers

Most general-purpose AI chatbots, including ChatGPT, set their minimum age at 13 and ask for parental permission up to 18. Once a child is 13, COPPA's protections in the US largely fall away, which is why the settings and parental controls on the tools your teen uses matter more.

Seven questions to ask any AI learning app

You can usually answer these from the privacy policy and help pages. If you can't, email the company. A good one will answer.

1. What exactly do you collect from my child? Look for a specific list, not "information you provide."
2. Do you store voice recordings? For how long?
3. What happens to homework photos after they're used?
4. Do you use my child's data to train AI models? If so, is it anonymized, and can you opt out?
5. Do you share data with third parties? Which ones, and why? AI apps typically use cloud and AI providers to operate. That's normal, but advertising partners are a different matter.
6. Is my child's data ever used for advertising? For a children's learning app, the answer should be no.
7. How do I see and delete my child's data? There should be a clear process and a timeframe.

Reading a privacy policy in five minutes

Privacy policies are long, but you can scan for the parts that matter. Use your browser's find function to search for these words:

• "children" or "COPPA": is there a section specifically about children? There should be.
• "voice" or "audio": how is it handled and how long is it kept?
• "train" or "improve": are conversations used to train AI?
• "advertising" or "marketing": is children's data used for ads?
• "third part": who else receives the data?
• "delete" and "retain": how long is data kept, and how do you remove it?

If the children's section is missing, vague or buried, treat that as an answer.

How PennPaper handles your child's data

We think parents should be able to find these answers in plain language, so here are ours.

• What we collect from your child: a first name so the tutor can address them, learning progress and session activity. We don't collect email addresses or location from children.
• Voice: processed in real time so your child can talk to the tutor. We don't build voice prints or other biometric identifiers, and voice isn't kept as a permanent recording. Sessions are transcribed so you can review them.
• Homework photos: used so the tutor can read the problem, then deleted from storage once they've been read. The lesson keeps the text of the question, not the image.
• AI training: we don't use identifiable student data to train AI models without explicit parental consent.
• Advertising: we never use children's information for behavioral advertising, and we never sell personal information.
• Your rights: you can review your child's sessions and progress in the parent dashboard, and request deletion at any time.

The full details are in our privacy policy. For the rest of what makes an AI tutor safe, see how to choose a safe AI tutor for kids.

Frequently asked questions

Do AI apps record my child's voice?

Many voice-based apps process your child's voice, and some store the recordings. Check the privacy policy for "voice" or "audio" to see how long recordings are kept and whether they're used for training.

Is my child's data used to train AI?

It depends on the service. Some general chatbots use conversations to improve models by default unless you opt out. Children's learning apps should not use identifiable children's data for training without parental consent.

What is COPPA?

The Children's Online Privacy Protection Act is a US law requiring online services aimed at children under 13 to get verifiable parental consent before collecting personal information, and to let parents review and delete it.

How do I delete my child's data from an app?

Look for a "delete account" or "privacy request" option in settings or the privacy policy. Under COPPA, services directed at children under 13 must honor parents' deletion requests.

The bottom line

AI learning apps need some data to work. They don't need all of it, forever, for any purpose. Ask what's collected, how long it's kept and whether it's used for anything other than teaching your child, and choose the app that answers clearly.

---

Sources: US FTC, Children's Online Privacy Protection Rule, 16 CFR Part 312, as amended 2025; US FTC (2023), Amazon Alexa settlement; UK Information Commissioner's Office, Age Appropriate Design Code; EU General Data Protection Regulation, Article 8