The South Korean government has unveiled a comprehensive plan on Nov. 6 to address the increasing issue of deepfake-facilitated sexual exploitation, mandating that platform companies remove flagged content within 24 hours of a request on Nov. 6. Under these new regulations, even possessing or viewing deepfake material out of curiosity can lead to significant penalties, such as imprisonment for up to three years or fines reaching 30 million won (about $21,468). Telegram and various other domestic and international platform providers, long recognized as channels for the distribution of sexually exploitative content, will now face heightened regulatory scrutiny.
The introduction of these measures follows the rise of deepfake-related crimes, which have become a major social concern. Critics have pointed out that current laws are inadequate, lacking sufficient punitive measures and regulatory frameworks to deter offenders. In response, the government is implementing extensive reforms to strengthen existing protocols.
The initiative focuses on “severe punishment” and “rapid removal of illegal content.” Those involved in creating or distributing deepfake content will face stiffer penalties, and individuals caught merely possessing or watching such material will also face harsh consequences. The maximum sentence for producing or disseminating deepfake sexual content has been raised to seven years, matching the penalty for the illegal recording of sexual content. Additionally, the government has increased penalties for threatening or coercing victims with child sexual abuse material to three and five years, respectively. The scope of undercover investigations by law enforcement will be broadened, allowing officers to infiltrate criminal networks and ensure the complete confiscation of proceeds from these crimes.
To better protect victims, the government will work closely with both domestic and international platform operators, such as Telegram and Naver, to guarantee that content is removed within 24 hours of being reported. Platforms are encouraged to block suspicious content preemptively and conduct reviews subsequently. Furthermore, an automated artificial intelligence (AI) system will be developed to detect deepfake content in real-time and send automated removal requests to platforms.
Oversight of platforms that serve as key channels for distributing deepfake content, notably Telegram, will be tightened. The government intends to establish additional communication pathways to ensure prompt actions and improved enforcement. Major platforms, including Naver and Meta (formerly Facebook), will face fines and corrective orders if they fail to meet obligations to prevent the dissemination of such content. A government official remarked, “Since launching a hotline with Telegram in September, we’ve managed to facilitate content removal within one hour to at least two days. We aim to accelerate this process further.”
According to police data from January to October, 506 individuals were apprehended for crimes involving deepfake content, with 81.2% being teenagers arrested. This high percentage, attributed to their familiarity with AI technologies, has sparked concerns that stricter punishments alone could result in a surge of young people gaining criminal records. To address this, the government plans to enhance education on preventing sexual crimes in schools and expand victim support services.
Despite these comprehensive measures, doubts remain about their long-term effectiveness. Perpetrators frequently bypass restrictions by creating new channels or migrating to different platforms, even after takedown requests. The current non-compliance rate with these requests is still around 30%, posing an ongoing challenge. A government representative stated, “We will issue recurring fines and corrective orders to international platforms that fail to comply. With strong governmental commitment and bipartisan legislative support, we believe these measures will yield meaningful results.”
