South Korea's Misinformation Law Faces First Test as 2,732 Reports Yield 27 Content Actions

Thousands of complaints, few removals and missing overseas data

South Korea's revised Information and Communications Network Act has produced 2,732 misinformation complaints across five domestic platforms, but just 27 reported instances of deletion or reduced visibility. Google, Meta, X and TikTok did not submit complaint and response figures requested by the regulator, leaving a substantial gap in the first examination of how the law is working.

The domestic figures cover different periods after the July 7 implementation date. Naver recorded 900 complaints through August 31, Kakao 895 through September 28, and DC Inside 829 through September 30. Nate Communications recorded 66 and Daum 42, also through September 30. Naver and Kakao took no content action on their complaints; DC Inside deleted 26 posts, while Nate restricted visibility in one case.

Those results put the practical operation of the law at the center of the debate. The government describes it as a framework for platform responsibility and compensation for victims, rather than a system allowing officials to declare posts false and remove them. Critics worry both about excessive moderation and about a framework that may be difficult to enforce.

The distinction matters: a complaint is not proof that information is false, and a platform decision is not a court judgment. The early figures reveal how companies responded to reports, not how many posts legally qualified as misinformation or how many victims suffered compensable harm.

What changed, and when?

The Broadcasting, Media and Communications Commission, chaired by Kim Jong-cheol, announced Cabinet approval of the amendments in its December 30, 2025, Cabinet approval statement. It identified three central changes: increased civil damages for qualifying publishers, moderation policies established by large platforms, and a transparency center supporting private factual verification work.

The official documents disagree on the implementation date. The December statement identified July 5 as the planned effective date, while the government's July 30, 2026, guidance says the law took effect on July 7. The subsequent platform complaint figures also use July 7 as their starting point. The reason for the two dates is not explained.

The sequence of announcements and reporting periods is:

• December 30, 2025: Cabinet approval was announced, with detailed eligibility rules still to be set through subordinate legislation.
• July 7, 2026: The implementation date identified in the government's subsequent guidance and complaint reporting.
• July 30, 2026: Government guidance described platform designations, publisher thresholds and complaint procedures.
• August 31: The cutoff for Naver's 900 complaints.
• September 28: The cutoff for Kakao's 895 complaints.
• September 30: The cutoff for Daum, Nate Communications and DC Inside.

The different cutoffs limit direct comparisons. Naver's figures cover a shorter period than those of several other services, so the totals should not be treated as a ranking of complaint rates over an identical interval.

Which platforms and messages are covered?

The government's July policy guidance sets the large platform threshold at an average of at least one million daily users. It names nine designated providers: Naver, Kakao, Nate, DC Inside, AXZ, Google, Meta, X and TikTok. These businesses must establish and operate their own policies and publish reports on their handling of illegal or manipulated false information.

The designation list and the complaint breakdown use different labels. The former names AXZ, while the latter separately lists Daum among the five domestic services. Readers should not add Daum to the nine provider list as though the two sets of names describe identical categories.

Private KakaoTalk conversations, direct messages and other private exchanges are outside the misinformation distribution rules, according to the government. The law does not grant authorities access to those conversations or powers to monitor them.

Public distribution is the dividing line, rather than the name of the app. KakaoTalk open chat may fall within the rules when an unspecified group of people can join and information circulates publicly. A private conversation and a publicly accessible discussion on the same service can therefore receive different treatment.

The government also says political criticism and personal opinions are not regulated simply because they challenge officials or offend someone. That assurance does not mean every publicly posted message is exempt: the legal assessment depends on its content, purpose and resulting harm.

What counts as manipulated false information?

The government describes a cumulative test, meaning all the required elements must be present. The publisher must know that information is wholly or partly false, or has been altered so that it can be mistaken for fact. There must also be an intention to cause harm or obtain an improper benefit, and an infringement of a legally protected interest, such as a person's reputation, property rights or the public interest.

This is narrower than treating every factual error as prohibited misinformation. An inaccurate statement alone does not establish the full test. Knowledge, purpose and harm also matter, including where a publisher earns money from advertising or views.

The official examples include a video that turns an unverified celebrity rumor into apparently factual content through sensational editing. Another is an interview whose statements are reordered, partly deleted or stripped of context to convey a meaning different from the original. Even in these examples, the required harmful intention or improper financial purpose must also be established.

Satire and parody are listed as exceptions, but the guidance gives a qualification. Material that appears satirical may still qualify if it meets the required conditions concerning harmful intent or improper gain and infringement of protected interests. Calling a post parody does not automatically settle its legal status.

Hate and discriminatory expression are treated separately from false information. The guidance covers direct incitement to violence or discrimination, or the serious promotion of hatred that substantially damages human dignity, based on attributes including race, nationality, region, sex, disability, age or social and financial status. Political criticism, satire and mere offensiveness are not sufficient by themselves; context and social effects must also be assessed.

How complaints and appeals work

Anyone can report suspected illegal or manipulated false information to a covered platform. A report must identify the content's location, such as its URL, explain the complaint and include supporting evidence. The reporting procedure also calls for the complainant's identity and contact details, rather than treating an unexplained allegation as a complete submission.

Filing a complaint does not automatically remove a post. Platforms review reports under their own policies and may delete or block content, reduce its visibility, suspend an account or take no action. The government says companies should conduct independent reviews, not simply count how many people complain.

Platforms must communicate their decisions and reasons to the complainant and publisher, with information about the appeal process. Both sides may appeal within six months of receiving notification. Unresolved disputes may be referred to the dispute mediation process at the Broadcasting, Media and Communications Review Commission, with court proceedings also available.

The legal roles remain distinct. A platform decides what to do under its service rules. Courts decide civil damages. The regulator separately decides whether a surcharge is warranted for repeated distribution, but only on the basis of a final court ruling. The government says it has no power under this framework to directly determine a post's falsity and order its deletion.

The December statement also says manipulated false information is excluded from the review commission's deliberative remit, unlike illegal information. That distinction should not be confused with the separate opportunity to seek dispute mediation.

Who faces damages of up to five times the loss?

The increased damages provisions target people who make a business of communicating facts or opinions to the public and meet specified revenue and audience thresholds. Under the July guidance, a publisher must have earned revenue, including advertising income, from at least three posts during the preceding three months.

The publisher must also have either at least 100,000 subscribers or an average of at least 100,000 monthly views for content posted during that three month period. The two audience tests are alternatives, not requirements that both must be satisfied. A monetized channel below the subscriber threshold can therefore qualify through its viewing figures.

Article 44-10 allows a court to award up to five times the actual loss when the applicable conditions are met. This is a maximum, not an automatic award for every false post or successful complaint. Liability and the amount depend on judicial findings, including the required intention, purpose and infringement of protected interests.

YouTube channels, news websites and media organizations' social accounts may fall within the rules if they meet the legal requirements. Public interest information, however, is excluded from increased damages. The December statement specifically identifies information about conduct covered by the Public Interest Whistleblower Protection Act, prohibited conduct under the Improper Solicitation and Graft Act, and comparable matters of public concern.

Article 44-11 adds safeguards against misuse by public figures, including candidates for public office and heads of public institutions. The statement describes an intermediate court ruling procedure, an obligation for public figures to disclose that ruling, and a mechanism for compensation in the opposite direction. These protections are intended to prevent damages claims from becoming a tool against information serving the public interest.

The billion won surcharge has a separate test

The maximum administrative surcharge of one billion won does not follow merely from uploading disputed information. The government says it applies to qualifying commercial publishers who distribute information at least twice after it has been recognized as misinformation in a final court judgment.

The publisher must also have earned revenue from at least three posts in the three months preceding the distribution. The official explanation of this surcharge does not repeat the 100,000 subscriber or viewing threshold used for increased civil damages. The two mechanisms should therefore not be treated as having identical eligibility tests.

The requirement for a final judgment separates the surcharge from an initial complaint or a platform's moderation decision. A platform may remove a post without that removal establishing grounds for the maximum financial penalty.

At the time of the early implementation reports, no related lawsuits or surcharge cases were reported. That is a description of the reporting period, not proof that no later cases exist or that the provisions cannot be used.

What the first platform figures actually show

The 1,837 complaints reported for Naver, Kakao and Daum are a subset of the 2,732 complaints across all five domestic services, not a conflicting total. Naver took no action on 900 reports, Kakao rejected 895, and Daum took no action on 41 of its 42 reports. The remaining Daum case was under review at the Korea Internet Self-Governance Organization, known as KISO.

DC Inside deleted 26 posts out of 829 complaints and rejected 803. Nate Communications restricted visibility in one of 66 cases. Across the five services, those 27 actions represent about 1% of the 2,732 complaints. DC Inside's deletion share was approximately 3.1%, compared with about 1.5% for Nate's visibility restrictions.

These calculations measure recorded responses to complaints, not the prevalence of misinformation. The figures do not disclose how many reports concerned the same content, why individual complaints were rejected, how many appeals followed or whether complainants eventually obtained compensation.

They also do not establish that platforms ignored valid claims. A low action rate could reflect complaints that failed the legal or service policy tests, restrictive interpretations by companies, or other factors. Without published reasons and appeal outcomes, the totals alone cannot distinguish between those explanations.

Transparency is the unresolved test

Google, Meta, X and TikTok did not respond to the regulator's request for complaint and action figures. Their absence prevents a comparison between domestic and overseas services and leaves the total volume of complaints across all nine designated providers unknown.

Lee Ju-hee, a Democratic Party lawmaker on the National Assembly committee overseeing science, technology, broadcasting and communications, called for closer checks on implementation. In a statement translated from Korean, she described the missing overseas figures as a barrier to assessing the law:

Without knowing the situation at major overseas platforms with many domestic users, we cannot fulfill the purpose of the law.

The enforcement limits were partly deliberate. The December government statement says punishment provisions associated with platform moderation policies were removed to preserve company autonomy and avoid excessive burdens. Later reporting also identified the absence of penalties for failing to publish the required reports as a weakness in securing compliance.

The proposed Information and Communications Service Transparency Center is a support body, not a single authority deciding truth. Its functions include supporting private factual verification activities, research, education, international cooperation and database development. Platforms are not required to accept private verification groups' conclusions.

The outstanding issues are concrete: whether moderation reasons become accessible, whether appeals correct mistakes, how coordinated reporting campaigns are handled, and whether overseas providers disclose comparable data. The early figures do not resolve those issues, and no next hearing, court deadline or publication date is identified.

Key Points

• Five domestic services recorded 2,732 complaints, with 26 deletions and one visibility restriction.
• Google, Meta, X and TikTok did not submit the requested complaint and response figures.
• Government guidance identifies July 7 as the effective date; the earlier Cabinet statement listed July 5.
• Private messages are excluded, but publicly accessible open chat may be covered.
• Misinformation requires knowledge of falsity or manipulation, harmful intent or improper gain, and infringement of protected interests.
• Qualifying commercial publishers can face court awards of up to five times the actual loss, with protections for public interest information.
• A surcharge of up to one billion won requires repeated distribution of information covered by a final court judgment.
• Complainants and publishers can appeal platform decisions within six months of notification.