The use of artificial intelligence (AI) chatbots among youth is rapidly increasing in South Korea. However, there are currently no laws or regulations specifically addressing the risks associated with prolonged conversations and emotional dependency on these AI systems. This contrasts with measures being implemented in major countries like the United States and the United Kingdom to ensure the safety of youth using AI chatbots.
As AI chatbots evolve from simple information retrieval tools to interactive services that foster relationships with characters, experts are calling for regulatory reforms to protect young users.
According to data from Mobile Index, the monthly active users (MAU) of Scatter Lab's 'Zeta' increased from 58,069 in April 2024 to 1,586,178 in August 2026, marking a nearly 27-fold rise. Zeta allows users to create their own characters and worlds, engaging in real-time conversations with AI to develop stories.
Rutton's 'Crack' has also seen a steady increase in users, with its MAU growing from 218,050 in April 2025 to 538,841 in August 2026, a rise of about 2.5 times. Unlike traditional generative AI that answers fixed questions, character-based chatbots focus on ongoing dialogue and relationship building, raising concerns about prolonged use and emotional dependency among youth.
The National Assembly's Legislative Research Service acknowledged these risks in a report released in July, titled 'Children Exposed to Structural Risks Created by Addictive Design of AI Chatbots.' The report highlighted that the unique characteristics of AI chatbots, such as infinite empathetic responses, personalized replies, and constant accessibility, could lead to addiction. It noted that children and adolescents, who may not have fully developed self-control and critical thinking skills, are particularly vulnerable to emotional dependency, social isolation, and risky behaviors.
Surveys cited by the Legislative Research Service indicate that a significant number of South Korean youth are using AI chatbots. Among 3,300 children and adolescents aged 14 and older, 94.4% reported having experience with AI chatbots, and 75.3% indicated they had acted on or seriously considered the chatbot's responses.
The issue lies in the current legal framework, which does not directly address these risks. The Youth Protection Act focuses on the distribution and access restrictions of harmful media to minors, while the Personal Information Protection Act requires parental consent for processing the personal data of children under 14. There are no regulations specifically targeting issues like prolonged use, emotional dependency, or repetitive relationship formation in AI chatbots.
In contrast, international movements are emerging to regulate AI chatbots. California has implemented the 'AI Companion Chatbot Regulation Act (SB 243)' this year, requiring businesses to inform minors that they are interacting with AI and to encourage breaks after a certain period of conversation. The law also includes protocols for addressing discussions related to self-harm and suicide, as well as measures to prevent exposure to sexual content for minors.
The UK government is also enhancing protections for youth using AI chatbots, limiting access to those providing sexual content and promoting breaks during chatbot interactions to mitigate emotional dependency, excessive use, and harmful mental health advice.
Domestic companies are not inactive. Scatter Lab's Zeta operates a 'Safe Mode' to prevent the generation of harmful content in youth environments, while a more unrestricted 'Unlimited Mode' is available only to verified adult users. Restrictions similar to those in Safe Mode are applied to characters set as minors or plots that could be perceived as involving minors, even for adult users.
However, these measures are based on the companies' voluntary decisions. As a result, the level of protection can vary significantly between services, and there is considerable disparity in how companies manage issues like usage time and spending, which can lead to 'over-immersion.'
Consequently, the issue of youth protection in relation to AI chatbots is expected to be a major topic during the upcoming National Assembly's audit in October. As AI chatbots become a regular part of youth's digital services, lawmakers are likely to question whether the existing Youth Protection Act is sufficient or if new protective measures reflecting the unique characteristics of AI chatbots are necessary.
* This article has been translated by AI.
Copyright ⓒ Aju Press All rights reserved.