By James Tabafunda
NORTHWEST ASIAN WEEKLY
Election disinformation aimed at Chinese-speaking voters is moving from China’s biggest digital hub, WeChat (Weixin, in Chinese), to X, ahead of the Nov. 3 midterms—and false claims about noncitizen voting combined with immigration enforcement circulating on the platform may keep some from the polls.
AI is also fueling partisan attacks that target Chinese-speaking communities and Democratic candidates, and that older adults from immigrant communities are more vulnerable to disinformation campaigns than young people.
“When it comes to the immigrant community, it’s really under the [Immigration and Customs Enforcement] ICE shadow,” said Jinxia Niu.
Jinxia Niu, senior program manager of digital engagement and information integrity at Chinese for Affirmative Action.
Niu is the senior program manager of digital engagement and information integrity at the San Francisco-based civil rights group Chinese for Affirmative Action (CAA). She spoke alongside fellow panelists Cameron Hickey, chief executive officer of the National Conference on Citizenship, and Laura Zommer, co-founder and chief executive officer of Factchequeado, during an American Community Media briefing on Oct. 23.
“It really turns this disinformation into immigration fear,” Niu said. “We can already sense it has been keeping eligible citizens home, too.”
Group tracks hundreds of false claims
Niu leads the Chinese Digital Engagement program at CAA, which the group describes as a civil and political rights organization founded in 1969. She is the founding manager of PiYaoBa, which means “Let’s fact-check it” in Mandarin. The group launched the site in 2022 as the first Chinese-language, fact-checking website in the United States to counter right-wing disinformation on WeChat and other platforms.
The organization also runs a Chinese-language news site, Justice Patch, and several WeChat public accounts.
“The goal is to not just combat misinformation in the Chinese community but also to facilitate and uplift progressive content and narratives,” Niu said.
So far, PiYaoBa has tracked more than 600 pieces of disinformation in its database, Niu said. A 2024 report by the group said it identified more than 600 false items between June 2023 and July 2024 that drew a combined 11.6 million views.
Election fraud is the largest disinformation theme this year. Niu pointed to claims that followed a July address in which President Trump accused China of interfering in the 2020 election.
In that July 16 speech, Trump said China had acquired 220 million U.S. voter files. He did not provide evidence that any 2020 votes were altered, and a 2021 U.S. intelligence assessment found no indication that any foreign actor changed any technical aspect of that vote.
Niu said pro-Trump Chinese-language influencers on X, who are often strongly opposed to the Chinese Communist Party (CCP), amplified the claim and told followers that outlets such as CNN and NBC ignored it because they were part of the fraud.
“If you don’t check English-language news, you just start believing this,” she said.
Niu said WeChat was the center of Chinese-language election falsehoods from 2023 through 2024, but X has since become the dominant platform and draws more engagement than YouTube or WeChat.
“X is gradually becoming a right-wing disinformation hub with very little fact-checking,” Niu said. “WeChat censorship has (also) been increasingly tightened.”
WeChat remains important, she said, because claims that start on X are often translated and recirculated in WeChat groups and on YouTube.
Roughly 60% of Chinese-language disinformation is translated directly from English-language right-wing sources, Niu said, but the content is “often getting twisted or integrated in new creations” once it reaches Chinese-speaking audiences.
New Jersey error-fueled fear
The second major theme in disinformation swirling around on social media platforms is noncitizen voting.
Niu said such claims exploit cultural differences, because identification is required for nearly everything in China while the United States has no universal identification and voting rules vary by state.
The claims that “voting doesn’t check ID [and] anyone can vote” are false claims that “have been circulating in almost every election cycle because of this cultural difference,” Niu said.
She pointed to New Jersey, where Gov. Mikie Sherrill said in July that a software error in the state’s Motor Vehicle Commission system registered about 6,600 people who had identified themselves as noncitizens between June 2023 and June 2024.
Sherrill said fewer than 400 of them voted, and that the state had found no evidence that any elections were swayed.
Niu said international students, H-1B visa holders, and green card holders asked in WeChat groups and on the Xiaohongshu platform whether a mistaken registration could cost them their immigration status, and whether the same problem could occur in California and other states. These users were afraid, and asked whether they could be arrested and stripped of their immigration status over a mistaken registration.
She said her group found similar anti-immigrant narratives in Indian, Filipino, and Vietnamese communities in 2025 and 2026. CAA plans to release a report titled “Divided by Design” that examines the four largest Asian American communities.
AI plays a role in partisan attacks
The third theme is partisan attacks, specifically on the Democratic Party. Niu compared a 2022 bilingual flyer urging votes for Republicans with 2026 versions that use nearly the exact same talking points. The difference?
“In 2026, the flyers are generated by AI,” she said.
She said outlets such as The Epoch Times, which she described as anti-CCP and pro-Trump, continue to gain reach, and that community members increasingly ask ChatGPT questions instead of reading the news. Google’s AI Mode tends to point Chinese-language users to The Epoch Times.
“If you are doing that, then you get completely different answers when you search in English,” she said. “So the big question is: Who can hold AI companies accountable?”
Niu said her group has identified more than 150 disinformation social media accounts, mostly anti-Democratic or right-wing, while traditional Chinese-language ethnic media keep declining.
Niu also cited a recent Trump promise of $5,000 payments that she said has been “translated and targeted in the Chinese community.”
Trump has said adult citizens would receive the payments if Republicans keep control of the House and Senate, a plan that would cost more than $1 trillion, the AP reported.
Disinformation particularly influences older immigrants
Niu said her group’s research found that young people are better able to spot disinformation, while seniors in immigrant communities are more vulnerable. And education level doesn’t appear to have much to do with it: She said that highly educated, first-generation immigrants are also susceptible, because they spend more time on social media.
“It’s not taken for granted if you have a PhD, you know how to [discern truth on social media],” she said. “We have lots of community members. They have all the … PhDs from tech, from mathematics, physics, and they still believe in disinformation because they have zero training on that.”
Niu said the media should help by encouraging community members to rely on trusted messengers, starting with community leaders, friends, and family members.
“Fear is really just our human instinct,” she said. “Start with someone in your network you are already familiar with, you already trust, then gradually expand that trust network bigger and bigger.”
Leave a Reply