The fine line between policing facts and silencing critics

New revisions to the Information and Communications Network Act, which will take effect in July, will hand Korea one of the most decisive legal tools yet devised against online falsehoods. For a country that has watched deepfakes and manipulated clips spread faster than fact-checkers can debunk them, this is surely a step worth welcoming. But it also deserves to be implemented with care, so that a sound principle does not curdle into overreach.

The revision targets influential online information producers, such as YouTubers with more than 100,000 subscribers or creators averaging over 100,000 monthly views. If such creators knowingly spread false or fabricated information and cause harm, they now face punitive financial damages of up to five times the loss incurred. Large platforms, defined as those with over a million daily users on average, must also establish formal reporting and response systems for disinformation.

The case for action is persuasive. The Hyundai Research Institute has estimated that fake news costs the Korean economy roughly 30 trillion won annually, about 1.9 percent of nominal gross domestic product. A 2024 survey by the Ministry of Science and ICT also found that more than 40 percent of respondents could not reliably distinguish real news from fake, a vulnerability more pronounced among older citizens less equipped to recognize artificial intelligence (AI)-generated content and deepfakes.

The enforcement gap has been just as troubling as the content itself. When the content in question sits on a global platform like YouTube or X, takedown requests can languish for weeks even under platform guidelines written specifically for election misinformation. Domestic platforms such as Naver and Kakao comply far more quickly, simply because they answer to local jurisdiction. The result is a two-speed system in which the most viral, most consequential falsehoods are often the hardest to remove in time, letting fabricated narratives settle into the public’s collective memory long before any correction arrives.

This is exactly where the new law aims to fill a gap. Victims of fabricated stories, whether ordinary citizens, public figures or institutions, have long had only the options of clearing up accusations personally or slow, expensive civil litigation that takes years and rarely restores what was lost, whether financially or in reputation. The punitive damages under the new revision gives them a faster, more direct avenue for redress, and the prospect of a public ruling against the offending party carries its own corrective weight, helping to undo some of the reputational harm even before any money changes hands.

Just as important, the law’s central value may lie in deterrence. A five-times punitive multiplier and the threat of a nine-figure fine can change the calculus for would-be disinformation peddlers before they hit publish, and encourages serious creators to better self-police their own claims. A media environment with creators with substantial influence on public discourse thinking twice before posting is, on balance, a healthier one.

Despite these positive implications, the law comes with its own complications, and getting the details right will matter. The first issue is definitional. What actually counts as “fake news” is inherently difficult to pin down, and any regulation that touches freedom of expression is going to be vague at the edges by nature. For instance, an article or an online video can be largely accurate while resting on one flawed source, or can make a sound argument from a partially mistaken premise. The new law appears to make an effort to exclude benign, unintentional errors from its scope, but drawing that line in practice could be substantially harder than stating the intent on paper.

Enforcement mechanics are worth considering as well. Take YouTube, for example: The online platform has utilized an AI classifier to scan videos and comments for content that looks like spam or fabrication, and an independent reviewer makes the final call. That two-step process is reasonably designed, but relies on an assumption that a clean line between true and false content always exists. Plenty of disputes are genuinely contestable, and an algorithm trained to flag suspicious patterns may not be necessarily equipped to referee a good-faith disagreement.

The platforms’ priority flagging system raises a similar concern. According to YouTube’s official guidelines, content flagged by government agencies or other reputable institutions is given priority for review and removal — a sensible rule since official bodies often have ready access to clear evidence. But this can disproportionately benefit those who already hold power. Consider a dispute between a whistleblower and the government agency they are accusing: a flag from the agency could see the whistleblower’s video removed quickly, while the agency continues to make its own case unchallenged.

That asymmetry is compounded by the harsh cost of an adverse ruling. Being adjudicated as a producer of fake news brings not only legal punishment but financial compensation to the alleged victim — up to five times the damage incurred. This is a penalty severe enough that it could chill speakers, especially those without institutional backing, from publishing material that turns out to be true but inconvenient for the powerful.

Many of these problems may be inherent to the platforms themselves, and no law can fundamentally provide a remedy for them.

The fairness and effectiveness of this law will ultimately be shaped less by the text of the decree than by the precedents Korean courts and regulators build around it in the years ahead. The principle behind it is sound and the intent to protect citizens from the very real economic and social damage of disinformation is one worth supporting.

The more complicated task ahead is to implement it with the same care that went into designing it, so that it strengthens public discourse rather than narrows it. As American civil rights activist Jesse Jackson once said, deliberation and debate are what stir the soul of democracy, and a law built to defend that soul should be applied in a way that keeps the debate alive and well.

Park Jin-wan is a nonresident fellow at the European Centre for North Korean Studies at the University of Vienna.