Taipei: Taiwan has flagged nearly 2,700 social media accounts to Meta that it suspects are part of a Chinese information warfare campaign aimed at influencing public opinion ahead of local elections next week, according to three government officials.
The accounts were identified on Threads, Meta’s text-based social media platform that has gained popularity in Taiwan. Taiwanese officials said the suspected accounts appeared to be part of coordinated efforts to spread divisive political narratives, undermine confidence in democracy and discourage people from voting.
Almost all the accounts were reportedly created within the past four months, with investigators identifying patterns suggesting they were opened in batches.
The November 28 elections for mayors and county chiefs are being viewed as an important political test for President Lai Ching-te‘s Democratic Progressive Party (DPP), ahead of Taiwan’s presidential elections scheduled for early 2028.
China considers the DPP a “separatist” force and refuses to engage with it. Taiwan, which is democratically governed, rejects Beijing’s claims over its territory.
Meta removes accounts flagged by Taiwan
A Meta spokesperson said the company had removed the accounts flagged by Taiwan’s government, adding that many had already been disabled by its own systems.
The spokesperson said none of the accounts identified as “low-quality fake accounts” had posted about election-related topics and that most had no followers.
“We will continue to take action against accounts that violate our policies,” the spokesperson said.
Taiwanese officials said they had confirmed that some of the accounts were taken down following their reports to Meta.
China’s Taiwan Affairs Office did not respond to a request for comment. Beijing has consistently rejected allegations that it interferes in Taiwan’s elections.
Taiwan’s National Security Council said China had used social media platforms and other channels to influence political discourse surrounding the island’s elections. Taipei said it would continue working with allies and responsible social media platforms to ensure that foreign interference did not affect the polls.
Suspected campaign sought to discourage voter participation
Three Taiwanese security officials investigating the activity said the suspected campaign formed part of a wider pattern of disinformation, amplification of divisive political narratives and attempts to undermine public trust in democratic institutions.
According to the officials, the suspected operations were not necessarily designed to promote specific candidates. Instead, they appeared aimed at intensifying political divisions and discouraging people from participating in elections.
“They don’t necessarily ask you to support certain candidates but they make people turn against each other, grow disgusted with democracy, choose not to vote, and they intensify election divisions,” one official told.
The officials spoke on condition of anonymity because of the sensitivity of the investigation.
Taiwanese authorities said the suspected activity was concentrated on Threads, which has become popular on the island. Officials said they had not identified similar accounts on Instagram.
The concerns come amid heightened scrutiny of foreign influence operations targeting democratic societies through social media, including the use of coordinated accounts and misleading content to shape political conversations.
Nearly 2,400 Threads accounts identified in one report
According to one internal Taiwanese government report, authorities flagged 2,400 Threads accounts as part of the suspected operation.
Of these accounts, 74% listed China as their location, around 25% did not disclose a location, while 1% listed Russia.
The report said the accounts were divided into 27 clusters, with similar Chinese names and different numerical IDs. Investigators believe the patterns suggested the accounts may have been created in batches.
“The assessment is that this batch of accounts was created in bulk by operators in China or Russia, likely with the aim of influencing Taiwan’s 2026 local elections,” the report said.
Officials said some accounts initially attempted to establish credibility by presenting themselves as ordinary Taiwanese users. They reportedly shared content about everyday interests, including fortune-telling, before shifting towards political messaging.
One official described how some accounts allegedly changed their messaging over time, initially appearing to be ordinary social media users before criticising the government and attacking candidates.
“Then one day, they start by criticising the government and attacking candidates, before saying they are disappointed with the ruling party and will therefore give up voting,” the official said.
Authorities believe such messaging could potentially discourage participation in the electoral process by encouraging distrust and political frustration.
AI-generated videos target DPP politicians
Taiwan also flagged around 260 accounts to Meta that officials said had circulated more than 2,150 AI-generated videos linked to the suspected campaign, according to the officials and another internal report.
The videos reportedly attracted 2.35 million interactions, including comments and shares. Most of the content targeted DPP politicians and policies.
Among those targeted was Puma Shen, the DPP’s prominent candidate for mayor of Taipei.
Some videos accused Shen of bullying when he was a student. Shen has denied the allegations.
The use of AI-generated videos and coordinated social media accounts has raised concerns among Taiwanese authorities about the potential spread of misleading information during election campaigns.
The investigation highlights Taipei’s efforts to identify and report suspected coordinated online activity to social media companies as the island prepares for its November 28 local elections.
