OpenAI Launches a Version of ChatGPT for Under-18s

OpenAI is activating default protections for users under 18 in a dedicated version of its chatbot, based on its internal principles and expert opinions. The platform includes a Study Mode that offers step-by-step guidance, reminders to avoid cheating, quizzes, and enhanced parental control options. This announcement comes as ChatGPT, launched at the end of 2022, had reached 900 million weekly users before the introduction of protective measures aimed at teenagers. The extent of the effectiveness of these measures against attempts to bypass them remains unknown until thorough evaluations have been conducted.

Default protections based on internal principles

OpenAI plans to enable age-appropriate protections by default in ChatGPT for Teens, with the stated goal of reducing exposure to harmful or inappropriate content for development. The company states that these safeguards are based on its Principles for users under 18, which it claims are informed by developmental science and experts. ChatGPT arrived in late 2022 and was credited with 900 million weekly users before the introduction of protections for minors. Meanwhile, OpenAI is already offering ChatGPT for Educators, a service aimed at educational institutions for managed access to AI along with support.

Study Mode guides learning and discourages cheating

The teen experience incorporates a Study Mode, designed to stimulate curiosity and problem-solving rather than seeking ready-made answers. According to OpenAI, this mode is intended to provide guiding questions and step-by-step support. When the user appears to be attempting to cheat, reminders of assignments will pop up, and the chatbot will direct them to Study Mode. The application will also include quizzes and learning visualizations, and controls will allow parents or guardians to set when Study Mode is activated by default. Announced on Monday, the service is presented as a response to AI-assisted academic cheating and is accompanied by a partnership with CodeAI to introduce teenagers to how AI works and how to direct or question it.

Uncertain effectiveness against bypassing and underlying litigation

The degree of difficulty for determined teenagers to bypass the new systems remains undetermined, especially since young users are often capable of circumventing parental controls and digital locks. In the absence of more rigorous testing of the teen mode, the actual extent of possible bypasses remains unknown. The rollout comes as plaintiffs in several cases have raised concerns about the lack of safeguards in AI chatbots, linking it to adolescent suicides and other mental health issues. OpenAI also indicates that parents will be able to rely on existing family tools to manage settings, receive safety notifications, and set Quiet Hours.