AI-generated fake news video spreads in South Korea
A YouTube video created with Google's Veo AI model depicted a fake news broadcast about lava erupting in Seoul, generated alarm on South Korean social media, and prompted discussions about the difficulty of distinguishing AI-generated from real video content.
What happened
A video created using Google's Veo AI model depicted a broadcast news scene in which an announcer reports lava erupting in Seoul and connects to a field reporter who reveals the lava behind him is fake and states 'I am AI.' Other figures in the video, including a student, entertainer, and businessman, each announce they are AI and warn viewers not to be deceived.
The video was created by YouTuber 'λΈκΉ λμμ΄λ' (Ddalgak Designer) to satirise people uncritically accepting content, inspired by news reports of romance scams involving deepfakes of Elon Musk's face. The video gained significant attention on social media in South Korea.
Reported impact
Netizens expressed both surprise at the video quality appearing nearly indistinguishable from reality and concern that advancing AI technology could be misused. Some noted that elderly people unfamiliar with AI tools may struggle to distinguish AI-generated from authentic content.
The creator noted that Veo makes it easy for anyone to generate videos and audio, and observed that awareness of the risks remains low despite the increasing accessibility of such tools.
Index analysis
Professor Choi Byung-ho of Korea University's AI Research Institute stated that if open-source AI video creation techniques are applied with malicious intent, determining truth becomes 'practically impossible,' and recommended that media and non-profit organisations run campaigns encouraging people to verify content sources and question everything.
Naver began offering an 'AI usage' labelling feature in June 2025 on its blog, cafe, TV and clip platforms to help users identify AI-generated content. South Korea's planned AI Basic Act, set to take effect next year, will require labelling of AI-generated films, dramas and other content services.
Governance controls
South Korea's upcoming AI Basic Act will mandate that generative AI content in films, dramas and other services be labelled as such. Naver introduced voluntary 'AI usage' labelling to allow creators to indicate whether AI was used in their content.
Classification
Sources and evidence
This record was researched and written by the Index, based on the following original reporting.