A bill amending Korea's Network Act (정보통신망법) advanced to committee review on 8 September 2026. As drafted, it would classify AI-generated synthetic "deepfake" images or video that depict an identifiable person in a way that causes sexual shame or damages reputation as statutory illegal information, require platforms to take emergency temporary measures upon receiving a report of such content, and expand the scope of a foreign provider's domestic agent to explicitly cover emergency-measure and technical/managerial compliance for this category. It applies to information and communications service providers generally, including foreign providers that operate through a domestic agent — but it is a bill under committee review, not a law in force, and no threshold, effective date, or penalty figure has been set.
What this means for you
- If you already have a Network Act domestic agent, note that its mandated scope would widen under this bill to include emergency temporary measures and technical/managerial obligations specifically for deepfake-type content — not only the agent's existing designated functions. This is a change to what the agent must be equipped to do, not a new designation trigger by itself.
- Review your content-report and takedown workflow now, ahead of any effective date. If your service allows user uploads of images, video, or synthetic media featuring real people, work out how an "emergency temporary measure" upon report would fit your existing moderation pipeline before it becomes a statutory obligation.
- Treat this as an early signal, not a deadline. The bill has no fixed threshold, effective date, or penalty yet — it is still at the committee-review stage and its scope can change before passage. Do not build compliance dates around it yet.
- Re-check after the next legislative step. Track whether the bill clears committee, since that is typically when thresholds, an effective date, and penalty amounts get fixed in Korean legislative practice.
Source: (no URL provided in this alert)