OpenAI is expanding parental controls in ChatGPT and positioning it as an educational tool for teens, building on its September launch of a system that lets parents link accounts to set time limits, turn off image-generation functionality and receive alerts if “something may be seriously wrong.” Now, parents will get notifications in cases where OpenAI deactivates their child’s account for making violent threats, though parents still cannot see the actual contents of chats. Starting in January, the firm’s technology began assessing whether users are minors to prevent them from seeing sensitive content involving graphic violence, self-harm or sexual role-playing. For accounts tagged as belonging to teens, ChatGPT will now default to suggesting learning-focused prompts and will remind heavy users to take breaks more often. Parents can also remotely turn on “study mode,” which walks users through problems step by step. OpenAI said it used input from educators and learning experts to design the system.