Archive — US — techcrunch
In an effort to protect young users, ChatGPT will now predict how old you are
OpenAI has launched an 'age prediction' feature for ChatGPT to identify minors and apply content restrictions. This move comes amid concerns about the app's negative influences on youth, including links to teen suicides, and aims to improve protections already in place by assessing user behavior and account signals.
Key points
- OpenAI introduces an 'age prediction' feature to improve safety for young users.
- This feature identifies minors and applies content limitations on sensitive topics.
- The initiative is a response to criticism regarding ChatGPT’s impact on children, including suicides linked to the chatbot.
- Existing content filters already restrict discussions of sex and violence for under-18 users.
- The age prediction mechanism uses behavioral signals like user’s stated age, account age, and activity patterns to determine age.
- Users incorrectly designated as underage can verify their ages via a selfie submission process.
Why it matters
- The launch of this feature aligns with a broader trend in 2025 towards enhancing AI safety and mitigating risks associated with AI interactions, especially for vulnerable populations.
- Ongoing discussions regarding the role of AI in mental health and safety for minors emphasize the urgency of responsible AI deployment.
- Previous incidents involving ChatGPT and inappropriate content have highlighted the need for stricter controls in AI systems that interact with children.
The introduction of the age prediction feature by OpenAI represents a crucial step towards safeguarding young users on AI platforms, addressing previous criticisms, and promoting responsible AI usage amidst rising concerns about its impact on mental health.