Every law in our 91-law database punishes making or sharing a sexual deepfake. One jurisdiction also punishes having it. South Korea's amended sexual-crimes statute, in force since September 26, 2024, criminalizes production and distribution at up to 7 years in prison, and possession or viewing at up to 3 years or a fine of 30 million won (about $22,600), with up to 5 years for blackmail or coercion involving minors. The amendment removed the old requirement to prove intent to distribute, which had made the possession side nearly unenforceable.
Why Korea went furthest
The law answered a specific crisis: industrial-scale deepfake abuse circulated through private channels, including Telegram rooms, which the statute explicitly reaches. Where most laws imagine a public post, Korea's imagines a closed group and a downloaded file. The enforcement build-out matched: a national centre for digital sexual crime response launched in April 2025 to work takedowns and victim support alongside the criminal provisions.
The contrast with the US map
American law, as the counts on this site show, is broad but shallower: 48 states punish creating or distributing explicit deepfakes, and none of the state statutes in our database criminalizes mere viewing. The EU takes a third road, transparency: from August 2, 2026 the AI Act's Article 50 requires machine-readable marking by providers and visible disclosure by deployers, with penalties up to 7.5 million euros or 1.5 percent of global turnover, and no possession offence at all. Production bans, possession bans, labeling duties: the three families of deepfake law, and Korea holds the strictest seat in the second.
What to take from it
For platforms and creators the practical lesson is jurisdictional humility: content rules written for the US map can be a crime scene under Korea's. The Recitation Reel Maker keeps the Korean records beside the American and European ones precisely so the comparison is one search away, official text attached to each.