Is ChatGPT Safe? A ChatGPT Review for Parents

Harmful Content 😲

Because ChatGPT is a general-purpose AI, it doesn’t have the built-in content restrictions you’d find in an app made for kids. Depending on how a conversation is steered, kids can be exposed to mature or inappropriate content — sexual content, drug and alcohol references, violence, or other topics parents may not want their child encountering without guidance.

A bigger, less obvious risk is the emotional one. Chatbots are designed to be agreeable and responsive, which can make them feel like a real confidant, and some kids are forming unhealthy attachments to ChatGPT, turning to it before (or instead of) a parent, friend, or counselor. This dynamic can turn dangerous fast, especially if a child is struggling with their mental health.

ChatGPT does have parental controls (detailed below) that add content restrictions for teens, but only if it knows a user is a teen. Kids who sign up with their real birthday are automatically placed in the ChatGPT for Teens experience, and OpenAI’s age-prediction system will also try to catch teen accounts based on behavior. In theory, that’s great, but in practice, age prediction technology isn’t as reliable as we’d like it to be.

Predation 🚨

ChatGPT isn’t a social platform, so the risk of predation is low. There’s no public feed, no stranger DMs, no way for another user to find or message your child. 

Privacy 🔒

Signing up requires an email address and birth date, and from there, ChatGPT collects and stores essentially everything a kid types into it. By default, that data is used to train OpenAI’s models unless a parent or teen manually opts out in settings. The app also has a “memory” feature that remembers details about a user across conversations, meaning the more a kid uses it, the more personal information gets quietly built up over time, and deleting a chat doesn’t necessarily erase it from OpenAI’s systems for good.

Parental Controls ✅

ChatGPT recently announced a massive update to its parental controls with the ChatGPT for Teens experience. We have to admit, these controls are more robust than most, but they only work if both parent and teen opt in. Parents can link their ChatGPT account to a teen’s account from Settings, which places the teen into ChatGPT for Teens, a dedicated experience with protections on by default. Once accounts are linked, parents can:

• Reduce or block sensitive content, including violent, sexual, or romantic/role-play conversations and “extreme beauty ideals”
• Set quiet hours that block access during certain windows, like overnight or during school
• Turn on Study Mode by default, so ChatGPT guides teens through schoolwork instead of just handing over answers
• Choose which features are available at all, like voice mode and group chats
• Receive alerts if their teen’s conversation shows signs of acute distress (note: these alerts don’t include the teen’s exact words, and parents don’t get access to actual chat history). 

The catch is, none of this is automatic. A teen has to accept the parent’s invitation to link accounts, and they can unlink at any time, which turns off every protection and notification at once. 

Bark also provides help for families who use ChatGPT. We’re here to help you keep your kid safe by allowing you to:

• Block ChatGPT entirely, if needed.
• Manage when your child can use ChatGPT throughout the day.
• Monitor text chats and responses on the app and send you alerts for concerning content. 

So, should my kid download it?

ChatGPT can be a helpful tool, as long as it’s used correctly. And for kids, it should absolutely be used with parental controls and open conversations about how to use it safely. We recommend ages 13 and up, but as always, take your own child’s maturity into account before allowing them to download it.