AI-Generated Fake Accounts Spread Political Messages in South Korea
Multiple social media accounts in South Korea used AI-generated and real-person images to spread conservative political messages including 'Yoon Again,' deceiving audiences. One account operator admitted the content was fabricated; civil society researchers documented a pattern of coordinated accounts using synthetic faces and stolen images of real women.
What happened
Social media accounts featuring AI-generated images of young women and images of real individuals posted political messages including 'Yoon Again' and anti-communist slogans on platforms including Instagram and Threads.
An account operator admitted the video was AI-generated after the content spread widely, stating 'I should have been honest from some point, but I missed the timing' and apologizing for deceiving many people.
Roh Moo-hyun Foundation director Hwang Hee-du documented cases where conservative supporters created accounts using AI-generated female faces or took photographs of actual female flight attendants and influencers without permission to spread political messages.
Reported impact
The AI-generated content reached a significant audience through social media algorithms, with Hwang noting that 'quite a lot' of people were deceived despite the videos being undetectable as fake to the human eye.
Hwang warned that the method creates 'fake public opinion' and represents 'cognitive warfare' attacking human perception, with potential consequences for democracy itself if left unchecked.
Index analysis
The incident demonstrates how AI-generated synthetic media and image manipulation can be deployed at scale through social platforms to create false political personas and manufacture artificial grassroots support.
The accounts leveraged algorithmic amplification and visual authenticity to reach users with low political engagement, illustrating how synthetic media can be weaponized for political influence campaigns.
Governance controls
The account operator's identity and intent were discovered through community investigation and documentation by civil society observers rather than through platform or regulatory action.
Classification
Sources and evidence
Sources not yet published
This record was researched and written by the Index from original reporting. Its sources are still being verified by an editor, and they are listed here once that check is complete. Until then the record carries no reviewer.