China-Based Hackers Impersonated US Voters With AI-generated Propaganda During 2022 Elections: Microsoft

A map of China through a magnifying glass on a computer screen showing binary digits on Jan. 2, 2014. Edgar Su/File Photo/Reuters

China-based hackers impersonated American voters online and used artificial intelligence (AI) to create and promote divisive online content during the 2022 midterm elections, according to a report by Microsoft.

The effort was part of a series of covert influence operations by the Chinese Communist Party (CCP) intended to mimic U.S. voters from across the political spectrum and create controversy along racial, economic, and ideological lines, according to the Sept. 7 report.

“Ahead of the 2022 U.S. midterms, Microsoft and industry partners observed CCP-affiliated social media accounts impersonating U.S. voters—new territory for CCP-affiliated [influence operations],” the report reads.

“These accounts posed as Americans across the political spectrum and responded to comments from authentic users.”

The report includes examples of visual content created by Chinese communist actors using artificial intelligence. These included images supporting “Black Lives Matter,” the false claim that “most black people are killed by police,” and more generic anti-American rhetoric.

The report states that the AI-generated content is more “eye-catching” than China’s previous attempts at overseas propaganda and will likely be improved upon and used against Americans in the future.

“We have observed China-affiliated actors leveraging AI-generated visual media in a broad campaign that largely focuses on politically divisive topics, such as gun violence, and denigrating U.S. political figures and symbols,” the report reads.

“We can expect China to continue to hone this technology over time, though it remains to be seen how and when it will deploy it at scale.”

World’s Largest Covert Influence Operation

Publication of the report comes less than two weeks after tech giant Meta purged thousands of accounts linked to Chinese law enforcement from its platforms.

Meta stated that the accounts were part of the largest known covert influence operation in the world and were engaged in spreading pro-Chinese communist and anti-U.S. propaganda.

Meta stated that the covert influence operation, dubbed “spamouflage,” was active on more than 50 platforms, including X (formerly known as Twitter), YouTube, TikTok, Reddit, Pinterest, Medium, Quora, and Vimeo.

“The network was run by geographically dispersed operators across China who appear to have been centrally provisioned with internet access and content directions,” the report reads.

“Taken together, we estimate Spamouflage to be the largest known cross-platform covert influence operation to date.”

Spamouflage also manufactured and attempted to distribute conspiracy theories apparently intended to undermine confidence in the United States and the rules-based international order. These included stories claiming that the United States started COVID-19 by shipping contaminated seafood to China, bombed the NordStream natural gas pipelines, and committed genocide.

Microsoft’s report also states that some 230 online influencers across a bevy of platforms promoted the CCP propaganda. Although these influencers claimed to be independent media, the report states that they were, in fact, employed by the CCP.

Undermining US From Within

The drastic uptick in malicious China-based cyber activity has been linked to the regime’s military and political doctrine of so-called mind dominance, a form of cognitive warfare designed to defeat the United States without entering an all-out war.

Before 2019, the CCP largely exploited the U.S. open media environment to promote positive views of China and communism. Its state-backed influencers used paid advertising and bot networks to earn favorable views abroad.

By 2022, however, a report (pdf) by cybersecurity firm Recorded Future found that the CCP had successfully pivoted into a new phase of influence operations, marked by the targeted messaging of well-defined audiences that were segmented based on granular demographic data, not unlike data used by global marketing and research agencies.

These new accounts, according to Recorded Future, likely receive guidance or material support from the CCP’s United Front Work Department, a powerful agency charged with overseeing global influence operations, and the regime’s top intelligence agency, the Ministry of State Security.

What’s more, the swift evolution in tactics and strategy could indicate that the regime is learning to conduct psychological warfare from a more senior partner such as Russia, whose own cyber tactics the new Chinese methods closely resemble.

By framing how Americans discuss the issues of the times and by polarizing Americans to the point that they’re unable to work with one another, the regime is, in part, working to ensure that the United States is incapable of meeting a crisis.

Perhaps nowhere is this ambition more clear than in the case of “Dragonbridge,” a malign influence operation conducted in support of the CCP and its goals.

Intelligence firm Mandiant, which uncovered the ongoing operation last year, found that the campaign sought to aggressively undermine U.S. interests by inflaming political tensions, discouraging Americans from voting, and claiming that the United States is covertly responsible for what’s, in actuality, CCP aggression.

The Dragonbridge operation did this by impersonating legitimate groups, much like this most recently discovered operation, and by plagiarizing and altering news articles and posing as U.S. citizens eager to criticize the U.S. record on race and social justice issues.