China is poised to influence South Korea’s upcoming general election by leveraging artificial intelligence technology, according to a report from a United States-based technology company on April 4. This development has sounded alarms for both voters and candidates preparing for the polls this Wednesday.
The Microsoft Threat Analysis Center (MTAC) has reported that a Chinese influence campaign is expected to target elections in South Korea, the United States, and India. The report highlights that South Korea’s general election, scheduled for April 10, will see over 30 million eligible voters casting their ballots. Additionally, India is gearing up for the world’s largest democratic election, with more than 960 million registered voters set to choose their parliament members in an election spanning two months, beginning April 19. The United States is also preparing for its presidential election, slated for November 5.
The MTAC has outlined that malicious actors are expected to create and distribute fabricated content online using AI technologies. This content is likely to promote narratives favorable to the Chinese government in each targeted country’s election. The report emphasizes, “While the current impact of such content on influencing audiences appears minimal, China’s growing experiments with enhancing memes, videos, and audio are set to continue and could become effective over time.”
According to the MTAC, cybercriminals linked to the Chinese government have been implicated in conducting influence campaigns aimed at Taiwanese elections. The findings reveal that between last December and January, these operatives launched a comprehensive assault utilizing AI-generated memes, audio clips, and fictitious news anchors. Notably, the report identifies a particularly active entity, dubbed “Storm-1376,” responsible for disseminating AI-generated audio clips featuring a presidential candidate during the election on January 13. The fabricated content misleadingly suggested that the candidate, who had actually withdrawn from the race in November, was endorsing another contender in the presidential election.
The Taiwanese cybersecurity firm, TeamT5, revealed in March that Chinese operatives had disseminated disinformation across international social media platforms in the lead-up to the election. According to their report, deepfake videos of a Taiwanese presidential candidate were circulated on TikTok by accounts suspected of being pro-China. Furthermore, TeamT5 uncovered a campaign against Taiwan’s President Tsai Ing-wen, involving a fake electronic book released last January. This publication, filled with political rumors about the president, was promoted through AI-generated avatars in videos on YouTube. The malicious content was not only shared widely on platforms like Facebook and X (formerly Twitter) but was also distributed via newswire services, amplifying its reach.
The report highlights, “China has been actively weaponizing social media platforms to disseminate disinformation and propaganda targeting Taiwan and other nations. Despite other major political parties having nominated their presidential candidates as early as May, we observed no substantial influence operations until November. Notably, the majority of the narratives within these campaigns were focused on criticizing pro-Taiwan candidates rather than endorsing pro-China contenders.”
