By Charlene Lin | Published on June 25, 2026

A Chinese network of fake accounts featuring attractive Asian women supposedly looking to marry Taiwanese men appears poised to target Taiwan’s November 2026 local elections, NewsGuard found. The network of 294 coordinated accounts on Meta’s Threads, which launched in May 2026, bears the hallmarks of a Chinese political influence campaign.  

Although NewsGuard could not definitively link the campaign to the Chinese government, it includes accounts that appear to belong to a network previously used by Chinese influence operations to target Taiwan. The accounts bear multiple signs of coordination, including the same posting schedule, similar account handles, and a consistent focus on targeting Taiwanese men. 

The purpose of such campaigns is to build credibility for the women, help them create network connections, and discover the behavioral history of the targeted men before the accounts are used to craft and amplify coordinated political messaging. 

NewsGuard found that 118 of the accounts — approximately 40 percent — follow the naming pattern associated with a network of malign actors that previously spread political narratives attacking the pro-Taiwan Democratic Progressive Party in February 2026. Others were previously identified by NewsGuard as pro-China automated bots. These accounts’ handles typically consist of an animal name followed by a six- to eight-digit number, such as @dinosaur.9281033, @giraffe.6804250, and @moose.6869374. Seventy of these accounts posted identical dating profiles.

As of June 24, 2026, the phony dating accounts have not posted political content. However, these accounts appear primed to do political mischief ahead of Taiwan’s local elections in November that could determine the island’s future leadership and policy toward China.

USING WOMEN AS BAIT

Using attractive women as a lure is a longstanding feature of Chinese influence operations. For example, in December 2025, NewsGuard identified a Chinese campaign that posted 115 videos on Threads and other platforms with fabricated quotes from attractive Japanese influencers to push the pro-China claim that the Japanese islands of Ryukyu belong to China. 

A previous influence operation posing as dating accounts attempted to interfere with Taiwan’s July and August 2025 recall elections that were aimed at unseating pro-China lawmakers, a vote China strongly opposed, the Taiwanese think tank Doublethink Lab reported. In that campaign, 91 posts by 51 China-linked accounts posing as Taiwanese women on dating sites posted the same political slogan, “I am Taiwanese, and I oppose pan-green,” a reference to the pro-Taiwan Democratic Progressive Party. 

In March 2026, social media research firm Graphika reported that it uncovered a Chinese-operated romance scam targeting older men in the U.S. with AI-enhanced dating profiles, often tricking them to pay fees or disclose sensitive personal data. That campaign did not incorporate political content.

LOCATION IS EVERYTHING … INCLUDING FOR SAVORY NOODLES

The Threads dating profiles use Traditional Chinese — a writing system unique to Taiwan —  and say the purported women are “currently in Taiwan.” However, location data indicates otherwise, another sign of inauthenticity. Most of the accounts do not share their location, but among those that do, most are based in China, NewsGuard found. A few list locations in other Asian countries, including Pakistan, India, and Bangladesh.

Moreover, the accounts often describe a supposed interest in Taiwan-specific hobbies or cuisine, yet the descriptions of Taiwan and its culture are frequently off-base, suggesting that the user is unfamiliar with local knowledge. For example, in a June 4, 2026, Threads post, user @bear.4905813 called Hsinchu rice noodles “sour and sweet,” when in fact, the dish is savory. 

NewsGuard also found that the profile images used by these accounts are often repurposed photos from legitimate Chinese and Taiwanese influencers, or were AI-generated.

TIME FOR A LUNCH BREAK

NewsGuard’s analysis found 90 instances of multiple accounts posting similar content in batches simultaneously, suggesting that the postings were automated in order to churn out a large volume more efficiently.

Indeed, the data points to a carefully controlled exercise. According to an analysis by Taiwanese open-source researcher Chia-Tzu Ho of 457 posts from the network over four days, the fake accounts mainly posted between 9 AM and 7 PM Beijing time, with a break from 12 PM to 1 PM. Those are typical Chinese work hours, including a lunch break, suggesting that the accounts are populated by workers in a Chinese office setting.

The timing of the emergence of the network also points to its political intentions. Earlier campaigns identified by NewsGuard and other organizations also typically launched half a year before the targeted election, with pre-positioned accounts. 

“Fake accounts usually start emerging roughly six months before the elections, and they would appear apolitical at first to attract and accumulate audiences,” Jerry Yu, analyst lead at Taiwanese think tank Doublethink Lab, told NewsGuard. “But when important events happen, these accounts are then used to manipulate public opinion on the basis of the audience they have established in the early stage.”

Meta’s press office did not return two June 2026 emails from NewsGuard seeking comment on the accounts.

Edited by Dina Contini and Eric Effron