September 20, 2025, 2:05 PM ET — OpenAI said it is strengthening protections for minors with ChatGPT teen controls, pairing an age-prediction system with new parental tools that link family accounts and tailor responses for under-18 users. The company frames the update as prioritizing teen safety over privacy and freedom when those principles conflict.
In posts this week, OpenAI outlined how ChatGPT teen controls will steer younger users to an “age-appropriate” experience and block graphic sexual content or self-harm guidance. The push follows months of scrutiny over AI and youth mental health; CEO Sam Altman acknowledged trade-offs while arguing that clearer safeguards are needed as teen usage grows.
OpenAI says the system will estimate whether someone is under 18 based on usage signals and, when uncertain, default to protections that restrict chats and features. Alongside ChatGPT teen controls, the company plans parental options that let adults link to a teen’s account, set limits such as blackout hours, disable memory and chat history, and receive alerts in rare cases when the system detects acute distress.
Under the hood, the age-prediction approach is meant to reduce reliance on self-attested birthdays. The ChatGPT teen controls model will route minors to a more limited assistant, and in some cases may seek stronger verification. OpenAI emphasizes that crisis pathways are guided by experts and could involve contacting parents or, in severe situations, authorities.
There are caveats. OpenAI concedes that parts of ChatGPT teen controls reduce privacy relative to adult accounts and that automated age estimation will make mistakes. The company also notes that teens can access alternative chatbots or open-source tools, limiting how far any single provider’s policy can reach without broader norms or regulation.
For industry and families, the move formalizes a “teen mode” baseline that others may emulate. If ChatGPT teen controls become standard, schools and parents could gain clearer levers for supervision, while platforms face pressure to align crisis protocols and content boundaries. The open questions are accuracy, appeals when systems misclassify age, and whether safeguards meaningfully reduce harm in the real world.
What’s Next for ChatGPT teen controls: OpenAI says parental controls will ship by month-end, with further tuning to age estimation and crisis response. Watch for pilots that test opt-in ID checks, clearer appeal flows for misclassification, and expanded educator settings.
- Ship linked parent–teen accounts and configurable limits
- Refine age estimation and add appeal/override options
- Standardize crisis-alert pathways with outside partners
By the Numbers
- 13+: stated minimum age for standard ChatGPT access
- 2 levers in teen mode: age-prediction routing + parental controls
- 1 crisis path: rare alerts to parents/authorities for acute distress
- 0 tolerance for explicit content in teen experiences
External Sources
- OpenAI — “Building towards age prediction” (age prediction, teen routing, crisis pathways). OpenAI
- OpenAI — “Teen safety, freedom, and privacy” (trade-offs & principles). OpenAI
- OpenAI — “Building more helpful ChatGPT experiences for everyone” (parental controls feature list). OpenAI
- The Guardian — report on age verification push after teen death. theguardian.com
- Business Insider — overview of age-detection plans and parental controls. Business Insider
- WIRED — analysis on balancing teen safety and privacy. WIRED
