OpenAI Introduces Parental Controls for Minor ChatGPT Users, Lets Parents Link Accounts

OpenAI finally rolled out new parental controls for ChatGPT on Monday. The San Francisco-based artificial intelligence (AI) firm first announced these safeguards for minors last month after the incident where a teenager committed suicide after confiding in the AI chatbot. Parental controls are part of a wider effort by the company to better protect children and teenagers from inappropriate responses from ChatGPT. OpenAI is also developing an AI-powered age detection system that predicts whether a user is under the age of 18 and applies age-appropriate settings.

ChatGPT Gets New Parental Controls

In a blog post, the AI giant announced the release of the new parental controls on ChatGPT. These will be available to all users globally on the web. OpenAI said the features will be expanded to mobile apps soon. There's no word on when the desktop apps will get these controls.

Parental controls can be set up by either the parent or the minor. One of the parties will need to send an invitation to the parent or guardian, or the teenager, and the other party will need to accept it. Once the minor accepts the account linking, the parent can manage the teen's settings from their own account. Notably, if the minor unlinks their account, the parent will be notified.

Once the linking is complete, parents will see new features to moderate the responses the minor gets while interacting with ChatGPT. These options are available via a control page. Parents can choose if they want to restrict one or all of the options.

The new controls include Quiet Hours, which let parents decide specific times when ChatGPT cannot be used. Parents can also turn off Voice Mode, Memory, and remove image generation capabilities. Additionally, parents can also opt out of model training to ensure their children's conversations aren't used to improve OpenAI's models.

By default, all linked teen accounts will also get improved safeguards. These accounts will see reduced graphic content, viral challenges, sexual, romantic or violent roleplay, and extreme beauty ideals. These settings can be turned off by the parents, but the minors cannot adjust these settings.

Apart from this, OpenAI has also built a new notification system for parents in circumstances when the minor might be facing a moment of emotional distress. If ChatGPT recognises potential signs that an underage user could be thinking of self-harm, it will flag the messages to a team of human reviewers, who can then flag them to parents via email, text message, and push alert on their phone. The only time the notification won't work is if the parent has opted out of the notification system.

On false alarms, the company says, “We are working with mental health and teen experts to design this because we want to get it right. 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.”

OpenAI says it is also working on establishing the process and circumstances in which the company will reach out to law enforcement or other emergency services. This would only be done if the system detects an imminent threat to life and parents cannot be reached, the post added. The AI firm highlighted that even in these situations, it will only share the information needed to protect a teen's safety.

Notably, the company claimed to have worked with experts, advocacy groups such as Common Sense Media, and policymakers to decide on the approach to create safeguards for minors.