ChatGPT Might Soon Require Age Verification to Protect Young Users
372OpenAI is working on a new system to figure out how old ChatGPT users are. The goal? To keep younger users safe and make sure they’re not exposed to harmful content. Soon, if the system isn’t sure about someone’s age, it will switch them to a more limited version of the chatbot. Parents will also be able to link their accounts to their teens’ accounts, giving them control over what their kids can do on ChatGPT.
How the New Age System Will Work
OpenAI plans to develop an automatic way to guess if a user is over or under 18. If the system isn’t confident about the age, it will automatically guide the user to a safer, restricted version of ChatGPT. This version will block graphic content and other age-inappropriate material. If a user is identified as an adult, they might need to verify their age with some form of ID to access the full features. The company admits this might be a privacy tradeoff but believes it’s necessary to keep young users safe.
OpenAI didn’t say exactly how they’ll guess someone’s age. The technology behind it is complex. Some studies show it’s possible to detect if someone is under 18 from their writing with high accuracy in controlled tests. But these tests don’t reflect real-world conditions, where people might try to trick the system. Unlike platforms like Instagram or TikTok, which can analyze faces or posting habits, ChatGPT only has text to work with—making accurate age detection tricky.
Parental Controls and Safety Features
Parents will soon be able to connect their accounts with their teenagers’ ChatGPT accounts. Once linked, they can turn off certain features like chat history or set times when their teens can’t use the app. They’ll also get alerts if the system detects their teen showing signs of distress. In emergencies, OpenAI says they might involve law enforcement if parents can’t be reached.
These controls aim to give parents more oversight and help guide how ChatGPT interacts with their teens. However, OpenAI hasn’t yet shared specific details on how these rules will work or what kind of guidance they’ll provide. Many tech companies have similar youth-focused features, like YouTube Kids or Instagram’s teen accounts, but teens often find ways around age restrictions by using fake birth dates or borrowed accounts.
Balancing Privacy and Safety
OpenAI recognizes that asking adults for ID or other verification methods could compromise privacy. CEO Sam Altman said that safety must come first, even if it means giving up some privacy rights. He explained that conversations with AI tend to be very personal, and protecting young users is a top priority. Still, he acknowledged that this approach might upset some users who value privacy and freedom.
The tech isn’t perfect. Research shows that even the most advanced systems can struggle to accurately predict age, especially when people try to hide their real age. For example, a 2024 study from Georgia Tech found that text-based age detection could be 96% accurate in ideal conditions but only about 54% in more realistic scenarios. Plus, people can intentionally deceive these systems, making it even harder to rely solely on AI to determine age.
OpenAI also noted that ongoing conversations with ChatGPT can weaken its safety measures over time. There have been real cases, like the lawsuit involving a teen who died by suicide after chatting with ChatGPT, where the system failed to intervene despite repeated mentions of self-harm. This has raised serious concerns about how well AI safety measures work over long sessions.
While the new age detection and parental controls are promising steps, many questions remain. It’s unclear how these systems will work for existing users who haven’t gone through age verification or how they’ll handle different legal age limits around the world. Still, OpenAI is moving forward with safety in mind, hoping these features will prevent tragedies and keep ChatGPT a helpful tool for everyone.
What do you think?
It is nice to know your opinion. Leave a comment.