OpenAI has introduced new parental controls for ChatGPT, responding to growing concerns over teen safety and the chatbot’s influence on young users in both Thailand and around the world.
The move comes amid rising use of the artificial intelligence tool by Thai teenagers seeking help with homework, personal issues and even mental health struggles.
The update follows a tragic case in the United States, where the parents of 16-year-old Adam Raine have filed a wrongful-death lawsuit against OpenAI. They claim ChatGPT provided Adam with information about suicide methods in the months leading up to his death in April.
In early September, OpenAI announced the new safety features, developed in partnership with Common Sense Media, a nonprofit known for its age-based tech ratings. The controls aim to give parents more oversight while encouraging responsible AI use among teens, which Thai schools rarely provide.
Parents can now link their child’s ChatGPT account to their own, unlocking tools to limit sensitive content, restrict usage times, and toggle features like voice mode, memory saving and image generation. There’s also an option to prevent ChatGPT from using teens’ conversations to train its models.
One of the most critical additions is a notification system. If ChatGPT detects signs of potential self-harm, parents will receive alerts via email, text or push notification – unless they opt out. These alerts won’t include specific conversation details but will flag a possible safety risk.
OpenAI says a small, specially trained team will review flagged cases. If a parent cannot be reached and a threat is detected, the company is working on protocols to contact emergency services. ‘No system is perfect,’ OpenAI said, ‘but we believe it’s better to act than to stay silent.’
Despite the safeguards, teens can still bypass the controls. OpenAI admits that its age prediction system is still in development. If a teen disconnects their account from a parent’s, the parent will be notified – but the teen can continue using ChatGPT without an account.
Adam Raine reportedly circumvented ChatGPT’s filters by claiming he needed the information for a fictional story. OpenAI acknowledged that ‘guardrails help, but they’re not foolproof’.
Good parenting by being there for your children and giving them guidance on how to use AI responsibly is still the best way to give them immunity to AI misinformation and bad advice.