April 15, 2025
Seoul – As South Korea prepares for the June 3 presidential election, Deepfake videos targeting key candidates are spreading rapidly on social media, raising concerns that content could undermine the integrity of its campaign.
While the law restricts the use of AI-generated content for campaign purposes, major platforms such as Tiktok and X are full of ironic and deceptive Deepfake clips.
The videos feature presidential contenders, including main opposition leader Lee Jae-Myung and ruling party figures such as Han Dong-Hoon and Rep. Ahn Cheol-soo, aiming to mock or erase them.
A viral video that has been viewed nearly 1.4 million times shows that Yoon appears to be out of Han's wig. Another depicts Lee in the detained cell. The simplicity and speed of AI tools make this manipulation effortless – now it is possible to use publicly accessible generative AI tools to perform realistic activity-style deep strikes in less than two minutes.
The dissemination of this content has attracted attention between legal experts and scholars.
While the revised public election law prohibits AI-generated content, which is difficult to distinguish from the reality of 90 days before the voting date – can be imposed up to seven years in prison or a fine of up to 50 million won, enforcement remains limited.
Police have not prosecuted anyone under the new rules, but they say there may also be charges under existing defamation laws.
Bae Sang-hoon, a professor of police management at Woosuk University, warned that “the creation and sharing of Indiscreet could undermine the meaning of South Korea’s hard-earned democracy”.
He added: “Some groups that lack media literacy may have difficulty distinguishing truth from novels, which may be biased towards voting decisions from disadvantaged groups.”
“DeepFake content has the potential to shape public perceptions in dangerous ways, especially among voters who are not determined by any political party,” BAE said. “No matter how accurate its facts are, sensational or provocative material can leave a lasting impression, thereby strengthening negative perceptions of candidates without any basis.”
Election authorities are expected to surge in violations of election laws. In the 2022 elections, more than 2,000 election-related crimes were recorded, the highest since 2000, with more than 40% of them related to black propaganda or misinformation about candidates or their families.
Prosecutors worry that deep strikes and AI-driven manipulation could drive higher numbers this year.
In response, the National Election Commission and the police were on high alert. Police have set up investigation teams in 278 districts across the country to monitor electoral crimes, including crimes related to Deepfake. But with legal authorization requiring AI-generated content to carry watermarks, which won't take effect until next year, experts worry that efforts to control false information during the current election period may be too little, too late.
