8 in 10 Korean social media users want AI content labels, citing distrust
gettyimagesbank
More than eight in 10 social media users in Korea said artificial intelligence (AI)-generated content and advertisements should be clearly labeled, reflecting growing discomfort over material that can appear artificial, misleading or difficult to verify.
In a survey released Saturday by research firm Opensurvey, 85.6 percent of 900 YouTube, Instagram and TikTok users said social media platforms should disclose whether content was made using AI. Among respondents, 54.1 percent said such labeling was “absolutely necessary.”
Support for disclosure was high across all three platforms: 88 percent among YouTube users, 85.3 percent among Instagram users and 83.3 percent among TikTok users.
Nearly half of respondents, 49.9 percent, said AI-generated content felt artificial or awkward. Another 39 percent said they questioned whether its information was factual, while 33.8 percent said undisclosed AI use made them feel deceived.
Only 11.4 percent described AI content as creative, 10.4 percent called it interesting and 9.6 percent said it felt new.
The skepticism also extended to advertising. Among the respondents who recalled a negative reaction, 13.2 percent said it stood out because it appeared to have been made with AI. Some 20.7 percent said ads felt exaggerated, while 18.7 percent said they were repeatedly exposed to such content.
Users said basic labeling alone was not enough. Some 37.8 percent wanted disclosures identifying which parts of a post used AI, while 33.7 percent said they wanted clearer notices for advertising or sponsorship.
“Users are deeply fatigued by the possibility of distorted information and deception behind AI’s convenience,” the report said, adding that platforms and creators risk distrust if they do not clearly disclose when and how AI is used.