Teens are increasingly turning to artificial intelligence for learning, creativity, homework help, and everyday curiosity. These tools unlock opportunity — but also raise safety considerations that deserve thoughtful, proactive guardrails.
Today’s young people are the first generation to grow up alongside AI — not as a novelty, but as a constant. That reality shapes how we build. We focus on making AI that is safe, age-appropriate, and trustworthy by default, grounded in protections and guided by principles that give families meaningful control.
How we’re building safer AI experiences for teens
OpenAI’s commitment to youth safety and wellbeing isn’t theoretical — it’s reflected in the features, safeguards, and resources we’ve built into our products and practices.
Age-appropriate behavior and protections
OpenAI uses a privacy-preserving age prediction system in ChatGPT to help identify when a user may be under 18 so that age-appropriate protections can automatically apply. When ChatGPT estimates a user is a teen, it applies additional safeguards to reduce exposure to content such as graphic violence, sexual role play, and self-harm depictions – designed with developmental safety in mind.
Learn about our age prediction approach:
https://openai.com/index/our-approach-to-age-prediction/
Parental controls and family oversight
We offer parental controls allowing parents to link their accounts with their teen’s ChatGPT account. These controls enable families to customize settings — restricting certain features and setting quiet hours — while preserving privacy.
Importantly, parental controls can also alert parents when ChatGPT detects signs of acute distress, including potential self-harm or suicidal ideation – helping families stay informed and intervene early when it matters most.
Learn about parental controls and family features:
https://openai.com/index/introducing-parental-controls/
Under-18 (U18) principles
ChatGPT’s Under-18 (U18) Model Spec outlines how we want our AI models to behave when we identify that a user is under 18. Informed by developmental science and expert input, these principles emphasize transparency, safety-first responses, and early intervention — helping ensure teens receive appropriate support.
Read about our Under-18 protections in the Model Spec:
https://openai.com/index/updating-model-spec-with-teen-protections/
Safety policies that protect youth
OpenAI’s Usage Policies prohibit harmful content across all ages, including related to self-harm, disordered-eating promotion, sexual exploitation, and violent wrongdoing. These rules help ensure our systems won’t be misused to produce materially harmful content — especially for younger users.
View OpenAI’s Usage Policies:
https://openai.com/policies/usage-policies/
Resources for families and educators
Technology alone isn’t enough. OpenAI provides AI literacy resources for parents, teens, and educators, offering guidance on thoughtful use, healthy boundaries, and supporting positive AI interactions at home and in learning environments.
Explore AI literacy resources:
https://openai.com/index/ai-literacy-resources-for-teens-and-parents/
Why strong rules still matter
In addition to robust safeguards, youth safety also requires clear, modern rules that set standards for the industry. AI systems are interactive and adaptive — not static websites — and their risks often emerge from context, interaction, and duration of use, not individual pieces of content.
That’s why OpenAI supports the Parents & Kids Safe AI Act, the first comprehensive, future-proof framework designed to protect children in the age of general-purpose AI that reflects today’s reality:
- Children are already using AI at scale
- Parents often lack visibility into how AI systems engage with their kids
- Existing online safety laws weren’t built with conversational, adaptive systems in mind
The framework takes a risk-based, privacy-respecting approach to protect children from foreseeable severe harms, gives parents meaningful tools, and encourages innovation aligned with child development and family wellbeing.
Making Safety the Default
AI will keep changing quickly. What won’t change is our commitment to protections that are clear and practical.
Safer Internet Day is an opportunity to listen and learn alongside families, educators, policymakers, and youth-safety leaders—and keep raising the bar on “safe by design” for general-purpose AI. We’re committed to iterating on safeguards as evidence evolves and expanding tools that help parents and the entire ecosystem make smarter choices.
Our goal is straightforward: help young people access the benefits of AI—learning, creativity, confidence—without asking families to guess what’s safe. That means responsible design, transparent controls, and a shared set of rules that prioritize teens’ safety and wellbeing.