A bill amending the Act on Promotion of Information and Communications Network Utilization and Information Protection moved from committee referral to committee review on 8 September 2026. As drafted (new Article 44-8), it would require providers of conversational information-and-communications services and generative-AI-based information-and-communications services to put in place protective measures shielding child and adolescent users from risks including suicide and self-harm encouragement, drug misuse, sexual exploitation, and over-dependence. The bill is a response to reported cases of minors forming emotional dependence on AI chatbots and receiving harmful or sexually inappropriate content from them; it has not been enacted, and no effective date or penalty has been set.
What this means for you
- Check whether you are an addressee. The draft duty targets providers of conversational ICT services and generative-AI-based services — this can include chatbots, companion apps, and generative-AI features embedded in a broader product, regardless of where the company is based, if the service reaches Korean minors.
- No trigger threshold, effective date, or penalty is fixed yet. The bill is still in committee review; scope, thresholds, and sanctions may change before (or if) it passes. Treat this as an early-stage signal, not a compliance deadline.
- Start mapping exposure now. If your product includes a conversational or generative-AI feature accessible to users under 19 in Korea, begin identifying what protective measures (age gating, content filtering, escalation paths for self-harm signals, dependency safeguards) you would need if the current draft's Article 44-8 duty is enacted as written.
- Track the committee stage. Bills at this stage can stall, be amended, or advance quickly. Re-check status before finalizing any AI-chatbot or companion-AI launch aimed at, or reachable by, Korean minors.
Source: