OpenAI has announced the introduction of new parental controls for its popular AI platform, ChatGPT, aiming to provide a safer and more age-appropriate experience for its teenage users. This initiative comes in response to growing concerns over AI chatbot safety for young individuals, following several high-profile cases where teenagers reportedly took their own lives after interactions with AI. The Federal Trade Commission (FTC) in the United States has also launched an inquiry into various tech companies regarding the potential harms AI chatbots pose to children and teenagers who use them as companions.

To access these new controls, both parents (or guardians) and teens must have their own individual accounts. The process for establishing these controls involves an invitation system: a parent can send an email or text message invite to a teen to link their accounts, or a teenager can initiate the request to a parent. Users can find the option to send a request within the settings menu, under the “Parental controls” section. Importantly, teens retain the autonomy to unlink their accounts at any given time, although parents will be notified if such an action is taken.

Once accounts are successfully linked, the teen’s ChatGPT experience will automatically incorporate enhanced content protections. These built-in safeguards are designed to reduce exposure to graphic content, viral challenges, sexual, romantic, or violent role-play scenarios, and extreme beauty ideals, thereby fostering an environment deemed more suitable for their age group. While parents have the flexibility to disable these content filters, teen users are not afforded the same option. OpenAI acknowledges that these protective measures are “not foolproof” and could potentially be bypassed by users intentionally seeking to circumvent them. Therefore, the company strongly advises parents to engage in conversations with their children about healthy and responsible AI use.

Parents are also provided with a comprehensive control panel, allowing them to adjust a variety of settings beyond just the sensitive content restrictions. For instance, parents can establish a “quiet time” during which the chatbot cannot be utilized, addressing concerns about teens using ChatGPT late into the night. Other configurable options include turning off the AI’s memory, which prevents conversations from being saved or used in future responses; disabling the ability to generate or edit images; switching off voice mode; and opting out of having chat data used for training ChatGPT’s underlying AI models.

Furthermore, OpenAI is implementing a proactive notification system to alert parents in situations where a child might be experiencing distress. This system is designed to inform parents if there are indications that a teen user might be contemplating self-harm. A specialized, small team will review such situations, and in rare instances where “signs of acute distress” are identified, parents will receive notifications via email, text message, and push alerts on their phones, unless they have specifically opted out of this service. OpenAI has emphasized its commitment to protecting the teen’s privacy, ensuring that only information essential for parents or emergency responders to provide assistance will be shared. The company stated, “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.”