One Tech Tip: OpenAI adds parental controls to ChatGPT for teen safety

LONDON (AP) — OpenAI said Monday it’s adding parental controls to ChatGPT that are designed to provide teen users of the popular platform with a safer and more “age-appropriate” experience.

The company is taking action after AI chatbot safety for young users has hit the headlines. The technology’s dangers have been recently highlighted by a number of cases in which teenagers took their lives after interacting with ChatGPT.

In the United States, the Federal Trade Commission has even opened an inquiry into several tech companies about the potential harms to children and teenagers who use their AI chatbots as companions.

In a blog post posted Monday, OpenAI outlined the new controls for parents. Here is a breakdown:

Getting started

The parental controls will be available to all users, but both parents and teens will need their own accounts to take advantage of them.

To get started, a parent or guardian needs to send an email or text message to invite a teen to connect their accounts. Or a teenager can send an invite to a parent. Users can send a request by going into the settings menu and then to the “Parental controls” section.

Teens can unlink their accounts at any time, but parents will be notified if they do.

Automatic safeguards

Once the accounts are linked, the teen account will get some built-in protections, OpenAI said.

Teen accounts will “automatically get additional content protections, including reduced graphic content, viral challenges, sexual, romantic or violent role-play, and extreme beauty ideals, to help keep their experience age-appropriate,” the company said.

Parents can choose to turn these filters off, but teen users don’t have the option.

OpenAI warns that such guardrails are “not foolproof and can be bypassed if someone is intentionally trying to get around them.” It advised parents to talk with their children about “healthy AI use.”

Adjusting settings

Parents are getting a control panel where they can adjust a range of settings as well as switch off the restrictions on sensitive content mentioned above.

For example, does your teen stay up way past bedtime to use ChatGPT? Parents can set a quiet time when the chatbot can’t be used.

Other settings include turning off the AI’s memory so conversations can’t be saved and won’t be used in future responses; turning off the ability to generate or edit images; turning off voice mode; and opting out of having chats used to train ChatGPT’s AI models.

Get notified

OpenAI is also being more proactive when it comes to letting parents know that their child might be in distress.

It’s setting up a new notification system to inform them when something might be “seriously wrong” and a teen user might be thinking about harming themselves.

A small team of specialists will review the situation and, in the rare case that there are “signs of acute distress,” they’ll notify parents by email, text message and push alert on their phone — unless the parent has opted out.

OpenAI said it will protect the teen’s privacy by only sharing the information needed for parents or emergency responders to provide help.

“No system is perfect, and we know we might sometimes raise an alarm when there isn’t real danger, but we think it’s better to act and alert a parent so they can step in than to stay silent,” the company said.

Is there a tech topic that you think needs explaining? Write to us at onetechtip@ap.org with your suggestions for future editions of One Tech Tip.