Taiwan flags nearly 2,700 suspected China-linked Threads accounts ahead of election
Accounts allegedly used mass registrations, political messaging and AI-generated videos to deepen divisions; Meta says it removed the reported accounts, but their actual impact remains under investigation.

Taiwanese government officials have reported nearly 2,700 Threads accounts to US social media company Meta over suspected links to a China-related online influence operation that may have sought to shape public opinion ahead of Taiwan’s Nov. 28 local elections. Officials said the accounts allegedly used mass registrations, political messaging and artificial intelligence-generated videos to deepen social divisions and erode public trust in democratic institutions.
According to a Nov. 20 report by Reuters, Taiwanese officials recently identified a large number of suspicious accounts on Threads and submitted the information to Meta for review. Government officials familiar with the matter and internal reports reviewed by Reuters said most of the accounts had been created within the previous four months. Some appeared in batches under similar names with different numerical identifiers.
One government report analyzed 2,400 of the reported accounts. About 74% listed China as their location, while roughly 25% did not disclose a location and about 1% listed Russia. The report said the accounts could be divided into 27 groups, showing signs of batch creation.
Taiwanese officials said some accounts initially posted general content about fortune-telling and everyday life to build the appearance of ordinary users. They later began criticizing the government and attacking political figures, while messages expressing political disillusionment encouraged people to give up voting. Officials said the activity did not necessarily directly support any particular candidate, but appeared aimed at widening political divisions and reducing voter participation.
In addition to text-based accounts, Taiwan also reported about 260 accounts to Meta for allegedly distributing more than 2,150 suspected AI-generated videos. According to the government report, the videos had generated about 2.35 million interactions in total. Most targeted Democratic Progressive Party politicians and policies, including Taipei mayoral candidate Shen Po-yang.
A Meta spokesperson told Reuters that the company had removed the accounts reported by the Taiwanese government, many of which had already been disabled by the platform’s systems before the report was submitted. Meta said the low-quality fake accounts had not posted election-related content and that most had no followers. The company said it would continue taking action against accounts that violate its policies.
China’s Taiwan Affairs Office did not respond to Reuters’ request for comment. Beijing has long denied accusations that it interferes in Taiwan’s elections. Taiwan’s national security officials said they would continue working with international partners and social media platforms to prevent external forces from affecting the election information environment.
With only days remaining before the Nov. 28 local elections, the case has again highlighted the challenges that social media platforms, AI-generated content and cross-border information operations may pose to democratic elections. Whether the accounts were part of an organized overseas operation, and the extent of their actual impact, remains under investigation.

