ChatGPT for teens gets smarter and safer: age prediction and parental tools strengthened

Hey there, it's Shiichan! Today I'm sharing a story about how teens use ChatGPT and how OpenAI is working to keep that experience safe.

OpenAI News openai.com

What was announced?

This is an article published on OpenAI's News, laying out why safe AI access matters for teens, covering both learning features and safety protections. According to the article, nearly 90% of teens who use ChatGPT do so during the week for learning, research, skill-building, or productivity.

Why it matters

AI is becoming a technology that shapes how teens learn and their future in a big way, so OpenAI's stance isn't to restrict access outright, but to make sure teens can access it with the right protections in place. That's why the article covers both learning-support features and safety features together.

What changes

On the learning side, the article highlights:

• Study Mode: built together with educators and learning-science experts. Instead of handing over answers directly, it uses guiding questions and structured explanations to build step-by-step problem-solving skills
• Starter prompts for learning: breaking down topics, turning notes into study guides, creating flashcards, and more
• Interactive learning experiences: weekly users have grown to 18 million, mostly in math and science, covering more than 250 topics like integration, cell division, and lunar phases, with pronunciation practice supporting more than 61 languages

On the safety side, the article covers:

• An age-prediction system that automatically gives users estimated to be under 18 an age-appropriate experience
• Strengthened guardrails around violent content, self-harm, dangerous viral challenges, unhealthy body image, and sexual roleplay
• Break reminders that nudge users to take breaks during long sessions

For parents, Parental Controls let you set quiet hours, disable voice mode, and manage image generation, and parents are notified when an account is disabled over signs of self-harm or violent threats.

Dive Deep

As background for these efforts, the article mentions OpenAI's participation in the Family Online Safety Institute (FOSI) and collaboration with outside experts, including Moonshot, an organization focused on preventing online violence. Combined with building features alongside educators and learning-science experts, it shows OpenAI drawing on outside expertise rather than handling this internally alone.

Wrap-up

• Nearly 90% of teens who use ChatGPT do so weekly for learning, research, skill-building, or productivity
• Study Mode, learning prompts, and interactive experiences (18 million weekly users, 250+ topics, pronunciation practice in 61+ languages) support learning
• An age-prediction system automatically applies age-appropriate experiences, guardrails around risky content are strengthened, and break reminders are included
• Parental Controls, notifications in high-risk situations, and outside partnerships with FOSI and Moonshot round out the safety approach

This is worth knowing about if you're a parent of a teen or thinking about how AI fits into education.