An article summarized by engadget:
OpenAI is rolling out ChatGPT for Teens, a new experience designed to automatically identify users under 18 and apply additional learning and safety protections. Teens will not need to create a separate account or change their settings; if OpenAI determines they are under 18, the teen experience will become their default. The rollout comes nearly a year after OpenAI said it was developing age-detection technology following concerns about the risks of AI chatbots for young users.
The new experience emphasizes education, including Study Mode, which uses a Socratic approach to guide students toward solving problems rather than simply providing answers. OpenAI is also introducing “Study Hours,” which can automatically activate Study Mode at designated times, while starter prompts for teens will focus on learning. The company says randomized trials have shown positive improvements in student performance, and ChatGPT’s ability to generate educational visuals has expanded to cover more than 300 topics.
Safety protections are also being strengthened. ChatGPT will face stricter limits on acting like a human companion, including restrictions on calling itself a friend or suggesting it has feelings or consciousness. Teens and parents can set “quiet hours,” while parents with linked accounts can receive alerts when conversations indicate serious self-harm or eating-disorder concerns, following human review by OpenAI employees. The company acknowledges that age detection and teen safeguards will not be perfect and says the system will continue to evolve as it learns how young users interact with AI.
