ChatGPT introduced new parental controls that allow parents to connect their accounts with their teens’ or children’s accounts. This gave parents the ability to manage and adjust features such as image generation, content filters and screen time.
Parents could prevent their teens’ conversations from being used for model training and received notifications if there were indications of possible self-harm. These tools aimed to increase teens’ and children’s safety, but with new features came new problems.
For some students, the idea of parental controls made sense. Annalise Wachowiak, sophomore theatre major, explained why.
“I think parental controls on ChatGPT and AI, which are a very unpredictable and uncontrollable system, make sense because it can be misused… so parental controls make sense for a problem that’s in a lot of media right now,” Wachowiak said.
For a quick tutorial and rundown, both parent and child need an account. The parent must invite the teen or child to connect accounts by email or text, and then the teen or child must accept the invitation. The feature was added to the settings tab, which many users could easily miss.
Among the main features was safety notifications. Parents usually could not see their teens’ or children’s chats unless the system flagged something and trained reviewers found signs of a serious safety issue.
Whether it’s helping students with homework, studying or simply answering quick questions, it made sense to introduce more tools to families for the safe use of AI. Sam Moffat, sophomore public relations major, agreed when asked about the new parental controls.
“I think it’s wise. ChatGPT is definitely a tool to be used and is used by many students here on campus. So, the introduction of parental controls is a great idea for parents to help their kids grow, as opposed to losing all critical thinking skills. ChatGPT should be an amplifier, not a substitute,” Moffat said.
Experts, however, expressed concern that parental controls could be easily bypassed. One of the easiest ways was for the teen or child to unlink accounts anytime. The parent would be notified if the accounts were no longer connected. Ultimately, it comes down to communication and honesty between parent and child.
Human reviewers who decided whether to trigger a potential parental notification were also involved when a notification is activated.
Lauren Haber Jonas, OpenAI’s head of youth well-being, addressed a similar question in an interview with Wired.
“We want to give parents enough information to take action and have a conversation with their teens while still maintaining some amount of teen privacy, because the content can also include other sensitive information,” Jonas said.
As with all parental controls, effectiveness and the need for oversight varied case by case. Once communication stopped, the need for control increased. Every family is different, and when using AI’s parental controls as one of the most powerful tools to emerge in recent years, the key remains communication.
