OpenAI Implements New Safety Features for Teen ChatGPT Users

OpenAI, the creator of the AI chatbot ChatGPT, is rolling out new safety features for users under the age of 18. Among the changes, the company is disabling the human-like voice response for new teen accounts to reduce the AI's perceived personhood.

Users identified as teenagers will also receive regular reminders to take breaks from the application, with prompts stating that ChatGPT is an AI and "it can wait." These measures are intended to encourage healthy usage patterns and reinforce real-world interactions, according to a company blog post. Teen accounts will be the default setting for users who identify as teenagers during signup or who are otherwise identified by OpenAI as being under 18. The company did not specify how it identifies underage users, though individuals under 13 are prohibited from using the product.

OpenAI stated that these changes are not a direct response to a specific incident where children believed ChatGPT to be sentient. However, there are growing societal concerns about users developing beliefs that AI tools are conscious. Last year, Mustafa Suleyman, head of AI at Microsoft, expressed worries about the societal impact of AI products that appear "seemingly conscious."