South Korea's revised Information and Communications Network Act (ICN Act) became enforceable on July 7, 2026 — eight months after the National Assembly passed it on December 24, 2025, and one week after the Korea Media and Communications Commission approved the final enforcement decree on June 29. Courts can now impose civil damages of up to five times proven losses on online content creators who deliberately spread false information for commercial or political gain. Platforms with more than one million daily active users face steep penalties for failing to police such content. The U.S. State Department has filed formal objections.
What the Law Actually Does
The enforcement decree defines its targets with numerical precision. A "major online information producer" subject to punitive liability is any individual who, in the preceding three months, posted at least three pieces of content and either accumulated over 100,000 subscribers or averaged more than 100,000 monthly views. Platforms with over one million daily active users must operate complaint and monitoring systems, process reports through the KMCC's transparency center, and remove confirmed unlawful content on demand. Crucially, platform CEOs face personal criminal prosecution for defying government removal orders — a provision with no real equivalent in the EU's Digital Services Act, to which South Korea's law is frequently compared.
The liability trigger is broader still. Courts may treat not just economic gains but "intangible benefits such as expanding social or political influence" as qualifying "unfair advantages." A political commentator who gains followers by sharing disputed information — without monetising at all — could theoretically face punitive damages. And a one-star delivery app review, as regulators have confirmed, falls under the same legal standard as an AI-generated deepfake of a politician.
The Case for the Law
Proponents have a genuine grievance worth taking seriously. South Korea's online environment has a documented problem with coordinated disinformation — AI-generated deepfakes used in election campaigns, fabricated health claims spread by high-follower influencers, and algorithmically amplified fabrications that outrun corrections. Existing legal tools — civil defamation claims, administrative complaints to the Korea Communications Commission — have been too slow to match viral misinformation. The law's architects argue that commercial incentives drive most harmful fabricated content, and that aligning liability with profit motive is a defensible policy lever. They are not wrong that a monetising YouTuber with 500,000 subscribers deliberately spreading fabricated medical information occupies a different legal category from an ordinary citizen's social media post.
The Proportionality Problem
The difficulty is that "false or manipulated" information has no self-defining boundary, and the ICN Act provides no independent pre-removal tribunal. Once a complaint is verified through the KMCC's transparency center, platforms must act or face liability. The incentive structure this creates is not subtle: platforms threatened with five-times damages and CEO criminal prosecution will over-remove. A legal standard that rational actors can only safely satisfy by erring heavily toward deletion is not proportionate regulation of harmful speech — it is a structural mechanism for chilling all contested speech.
The International Press Institute condemned the bill immediately after its December 2025 passage, warning that vague language around "public harm" and the complete absence of press freedom carve-outs create serious risk that government officials and corporations will weaponise the law against journalists. A National Assembly petition opposing the law gathered over 140,000 signatures in the weeks before July implementation — a significant figure for a petition process that ordinarily attracts little public attention.
A DSA Without the Safeguards
The comparison to the EU Digital Services Act reveals the structural gap. The DSA imposes removal obligations and transparency requirements on very large platforms — but it explicitly prohibits mandatory general monitoring, preserves safe harbor principles for hosting intermediaries, and creates independent Digital Services Coordinators with structured appeals processes. South Korea's ICN Act does none of these things. It permits mandatory general monitoring, exposes platforms to criminal CEO liability, and grants the KMCC broad enforcement discretion without comparable procedural safeguards for accused creators or the platforms they use. Calling this framework DSA-inspired obscures the precise structural elements that would have made it proportionate.
Washington's Objection
The State Department's concerns go beyond protecting U.S. platform revenues in Korea. Officials warned publicly that the law could enable viewpoint-based censorship beyond its stated targeting of deepfakes — precisely because "false or manipulated information" as a government-directed removal category has no counterpart in U.S. platform liability law under Section 230. Seoul's National Security Adviser confirmed bilateral talks occurred before passage and acknowledged that American perspectives were incorporated into the final text. That the State Department continues to object publicly suggests the accommodations were minimal. Google, Meta, Naver, and Kakao all face direct compliance obligations under the one-million-user threshold — and for U.S. platforms, the CEO criminal liability clause is without precedent in any jurisdiction where they currently operate under treaty-level relations with Washington.
The Chilling Effect Is Structural
South Korea's free speech jurisprudence already permits criminal defamation for truthful statements made with intent to harm reputation — a feature that has long distinguished it from most liberal democracies. The ICN Act builds on that tradition rather than departing from it. The danger for the innovation ecosystem is not limited to press freedom: platforms uncertain where the "false information" line falls will make conservative, defensive removal decisions that suppress legitimate journalism, political commentary, and ordinary consumer reviews alike. Jurisdictions with the lowest liability tolerance tend to define the global content moderation floor. If South Korea's punitive damages model for contested speech spreads — and regulators in other markets will cite it as precedent — the internet gets smaller. That contraction will not stop at the Korean border.