Police officer arrested over AI-generated sexual deepfakes

A police officer at Dongdaemun Police Station in Seoul was arrested Sept. 16 for allegedly generating and distributing sexually exploitative deepfakes through Telegram. Police said he used AI to create more than 400 explicit deepfakes of people he knew and continued the activity for more than two years while serving on the force. Investigators also found thousands of illicit images and videos in his possession.

Police said investigators found about 5,600 images and videos in his possession, including roughly 2,600 deepfakes, 1,600 child and adolescent sexual abuse materials and 1,300 illegally filmed videos.

The case began in a Busan Metropolitan Police Agency investigation into a Telegram chat room where sexually explicit deepfakes were allegedly being shared.

Police obtained court approval in December last year to run an undercover investigation of the chat room, and a concealed officer tracked its activity for several months before identifying the suspect.

The officer joined the police force in April 2024, after the alleged offenses had already begun, and police said he had previously been investigated for a similar offense before joining the force.

Published Oct 4, 2026 4:29 pm KST

Alleged offenses began 2 years before joining police, continued while on duty

A police officer has been arrested last month for allegedly generating sexually exploitative deepfakes of people he knew with the use of artificial intelligence (AI) and sharing them on Telegram, police said Sunday.

According to police, the officer, assigned to Dongdaemun Police Station in Seoul, was arrested Sept. 16 on charges including possessing sexual abuse material and sexually explicit deepfakes, as well as creating and distributing such content.

Investigators found some 5,600 images and videos in his possession, including about 2,600 deepfakes, 1,600 pieces of child and adolescent sexual abuse material and 1,300 illegally filmed videos. Police also suspect that he personally created and distributed some of the illicit content.

The case came to light during an investigation by the Busan Metropolitan Police Agency into a Telegram chat room where sexually explicit deepfakes were allegedly being shared.

Police obtained court approval in December last year to conduct an undercover investigation of the chat room. An investigator entered the group under a concealed identity and tracked its activity for several months before identifying the officer as a suspected distributor.

Authorities later confirmed that the suspect was an active police officer assigned to the traffic division of the Dongdaemun Police Station in Seoul.

Undercover investigations of this kind were introduced in September 2021 through revisions to the law on the protection of children and juveniles against sexual abuse, following a series of major digital sex crime cases that drew nationwide outrage.

Police searched and seized the officer’s phone and other belongings in August. While the warrant initially cited the alleged distribution of just three deepfake images, a forensic analysis uncovered a far larger volume of illicit material.

Police allege that between March 2022 and July this year, the officer used AI applications to create more than 400 sexually explicit deepfakes depicting people he knew. He is also accused of secretly filming women and distributing dozens of deepfake images through Telegram chat rooms between August last year and April.

The officer joined the police force in April 2024, about two years after his alleged offenses began, and is accused of continuing them for more than two years while serving as a police officer.

He had also previously been investigated for a similar offense before joining the police and received a suspended indictment, according to reports.

The officer reportedly confessed to the allegations after the police investigation intensified.

首尔东大门警察署一名警员于9月16日被捕，他涉嫌通过Telegram制作和传播色情深度伪造视频。警方称，该警员利用人工智能技术制作了400多张其熟人的露骨深度伪造视频，并在任职期间持续从事此类活动两年多。调查人员还在其住所发现了数千张非法图片和视频。

警方表示，调查人员在他身上发现了约 5600 张图片和视频，其中包括约 2600 个深度伪造视频、1600 个儿童和青少年性虐待材料以及 1300 个非法拍摄的视频。

该案源于釜山地方警察厅对一个 Telegram 聊天室的调查，据称该聊天室分享了露骨的深度伪造视频。

去年 12 月，警方获得法院批准，对该聊天室进行卧底调查。一名便衣警察跟踪了该聊天室的活动数月，最终确定了嫌疑人。

该警官于 2024 年 4 月加入警队，当时涉嫌犯罪行为已经开始。警方表示，他在加入警队之前就曾因类似罪行接受过调查。

发布于2026年10月4日下午4:29（韩国标准时间）

据称，犯罪行为始于加入警队前两年，并在执勤期间继续发生。

警方周日表示，一名警官上个月因涉嫌利用人工智能 (AI) 制作其认识的人的性剥削性深度伪造视频并在 Telegram 上分享而被捕。

据警方称，这名隶属于首尔东大门警察局的警官于9月16日被捕，罪名包括持有性虐待材料和性露骨的深度伪造视频，以及制作和传播此类内容。

调查人员在他身上发现了约5600张图片和视频，其中包括约2600张深度伪造图片、1600份儿童和青少年性虐待材料以及1300段非法拍摄的视频。警方还怀疑他本人制作并传播了部分非法内容。

该案件是在釜山地方警察厅调查一个 Telegram 聊天室时曝光的，据称该聊天室中有人分享露骨的深度伪造视频。

去年12月，警方获得法院批准，对该聊天室进行卧底调查。一名调查员以匿名身份进入该群组，跟踪其活动数月后，确认其中一名警官为涉嫌分发者。

当局随后证实，嫌疑人是首尔东大门警察署交通部门的一名在职警官。

2021 年 9 月，在一系列引起全国公愤的重大数字性犯罪案件之后，通过对保护儿童和青少年免受性虐待的法律进行修订，引入了此类卧底调查。

警方于8月搜查并扣押了该警官的手机和其他物品。虽然搜查令最初仅指控其传播了三张深度伪造图像，但法医分析发现其非法材料的数量远不止于此。

警方指控，这名警官在2022年3月至今年7月期间，利用人工智能应用程序制作了400多张包含性暗示内容的深度伪造视频，视频中的人物都是他认识的人。他还被指控在去年8月至今年4月期间，秘密拍摄女性视频，并通过Telegram聊天室散布数十张深度伪造图像。

该警官于 2024 年 4 月加入警队，大约在他被指控的犯罪行为开始两年后，并被指控在担任警官期间继续犯罪两年多。

据报道，他在加入警队之前也曾因类似罪行接受调查，并被缓刑起诉。

据报道，在警方调查深入后，该警官承认了这些指控。