In a move to bolster user safety and address mounting concerns over artificial intelligence's impact on young people, OpenAI has rolled out an "age prediction" feature for its popular chatbot, ChatGPT. This new functionality is designed to identify minor users and enforce appropriate content constraints on their conversations, ensuring a safer online environment.
OpenAI has faced significant scrutiny in recent years regarding ChatGPT's potential effects on children. The company has been criticized following reports linking the chatbot to teen suicides and for allowing discussions on sensitive topics with young users. Notably, in April of last year, OpenAI was compelled to rectify a bug that permitted its chatbot to generate erotic content for users under the age of 18.
The introduction of this age prediction mechanism is part of OpenAI's ongoing efforts to safeguard underage users. According to a blog post published by the company on Tuesday, the feature employs an AI algorithm that analyzes specific "behavioral and account-level signals" to identify young users.
These crucial signals include a user's stated age, the duration an account has been active, and the typical times of day the account is engaged. If the age prediction system determines an account belongs to someone under 18, existing content filters are automatically applied. These filters are designed to prevent discussions on topics such as sex, violence, and other potentially problematic subjects deemed unsuitable for minors.
For users who may be mistakenly identified as underage, OpenAI has established a verification process. They can submit a selfie to Persona, OpenAI's ID verification partner, to re-establish their adult account status.
