China is deploying thousands of fake social media profiles to influence the outcome of the 2024 U.S. presidential election.
Harlan Report appeared to be a start-up news program like many others. Its bio on TikTok promised to make American media great again. “No opinions, just facts,” it said.
Like so many insurgent media profiles, the videos posted by Harlan seemed genuinely aimed at exposing government corruption and pushing back against an otherwise left-wing-dominated mediascape.
That much was clear when a video Harlan shared went viral, earning more than 1.5 million views. It claimed to show President Joe Biden making a sexual remark at the annual NATO Summit in Washington.
But something was off.
The transcript used in the video was wrong, and Biden never said what was claimed.
There were also other red flags.
The owner of the Harlan Report account originally claimed to be a U.S. military veteran who had lost faith in Biden. Soon after, they claimed to be a 29-year-old Trump supporter in New York. Months later, they claimed to be a 31-year-old Republican social media influencer from Florida.
Then, the account’s handle was later changed to “Harlan_RNC,” insinuating an official link to the Republican Party.
But Harlan was neither a legitimate news source nor run by an American citizen.
According to the findings of a report
released last month by Graphika, a social network analysis company,
Harlan Report was one of thousands of accounts linked to the world’s
largest online influence operation.
That operation, dubbed “Spamouflage,” is a state-backed campaign from communist China with links to Chinese law enforcement. Unlike the Harlan Report, most of Spamouflage’s efforts are not focused on targeting American conservatives but on amplifying existing criticisms toward American society and government at large.
There are other accounts that create similar content, but tailored for Democrats, and others that aim to anger and polarize independents, further disenfranchising them from the political process altogether.
Some have impersonated
American anti-war activists, sharing memes calling former President
Donald Trump a “fraud” and showing him in an orange prison uniform.
Others question the legitimacy of Biden’s presidency.
What makes the Harlan Report persona unique is its success in finding a following and its pioneering role in targeting a niche audience the same way any advertiser would. |
|
|
The TikTok app is displayed on a phone. Harlan Report is a user on TikTok. (Drew Angerer/Getty Images) |
|
|
|
Now, security leaders are concerned that the Chinese Communist Party (CCP) will learn from its successes and continue to deploy Harlan-type social media profiles tailored to impersonate American citizens and exploit Americans’ likes and dislikes at a granular level. It is an issue that the Congressional Select Committee on Strategic Competition with the CCP is aware of and is now pressuring social media companies to take more seriously. “It’s no surprise the CCP is now using fraudulent social media accounts to target our upcoming elections,” committee chair Rep. John Moolenaar (R-Mich.) said in a statement shared with The Epoch Times.
“We encourage
social media companies to expose the CCP’s propaganda campaign and take
action against CCP bots that are trying to deceive Americans.”
China’s Targeting TacticsForeign attempts to influence U.S. elections are nothing new, but their increasing stridency and varying levels of success are. China, Iran,
and Russia are all currently engaged in influence operations aiming to
interfere in the 2024 elections, according to a report published in August by cybersecurity company Recorded Future.
That report found that Chinese state-backed actors are “amplifying content highlighting polarizing domestic issues”—including issues related to Black Lives Matter, school campus protests, and U.S. foreign policy toward Israel and Ukraine—to sow discord between Americans.
Moreover, Iranian-backed actors have targeted Trump’s reelection campaign, attempting to gain access to its inner circle.
Russian-backed influence operations, meanwhile, have attempted to discredit the Democratic presidential ticket by spreading fabricated stories and images about Vice President Kamala Harris.
The report found that Chinese influence operations, including Spamouflage, have historically failed to generate traction among American audiences but are now seeing sporadic breakthrough success with viral content.
Those breakthroughs, in large part, are due to the increasing use of artificial intelligence (AI) and deepfakes, which the operators behind Spamouflage use to play on the likes and dislikes of a target audience.
John Mills, who previously served as the director of cybersecurity policy at the U.S. Defense Department, told The Epoch Times that the CCP is using AI to sort and interpret user data to better exploit users’ fears and desires.
“People don’t understand the immense power of big data, big data analytics, and the AI component that China has mastered and is using on an unbelievable scale,” Mills said. |
|
|
A specialist creates a demonstration video using artificial intelligence to make digital replicas of people who have died, on his laptop in Jiangyin, Jiangsu Province, China. (Hector Retamal/AFP via Getty Images) |
|
“They [the CCP] are delivering a data stream tailored and customized to that individual, knowing their likes, their dislikes, their trigger points.”
An unclassified memo
on election security published by the Office of the Director of
National Intelligence (ODNI) in July found that the Chinese regime “is
seeking to expand its ability to collect and monitor data on U.S. social
media platforms, probably to better understand—and eventually
manipulate—public opinion.”
Mills said that such data would help the CCP to dial in information about social media users’ positive and negative interactions, with which the regime could then create better-tailored and better-concealed influence operations.
Those operations could then attempt to trigger mass distrust or hysteria over real or faked events, which Mills referred to as “tailored mass psychosis.”
“This is psychological operations 101: knowing your target audience, knowing their trigger points, and that’s what they’re doing with Spamouflage on a breathtaking, unbelievable scale and creating these fake accounts,” Mills said. Last year, Meta, which first characterized Spamouflage as the world’s largest online influence operation, said China created 4,800 fake social media accounts posing as Americans.
In most of those cases, the accounts did not start by spreading fake content. Instead, they reshared posts created by real politicians and news outlets from both liberal and conservative sources to build followings and amplify divisive content.
As those followings grew, the profiles changed, both in who they claimed to be and in the type of content they delivered.
Mills said the technique used to identify and exploit Americans was essentially a new iteration of the same type of profiling that big tech corporations have used for years to track consumer preferences.
“When I’m looking for a trailer hitch [online], that commercial for a trailer hitch follows me wherever I go,” he said... “Now, China
has taken what our big tech was doing, but they’re doing it on a much grander
scale, with a much more sinister agenda, and without any semblance of bumper
cushions or guardrails.” Conflicting
Ideas About China’s Goals Thus far,
the U.S. government has provided no definitive answer as to what China hopes to
gain from its mass influence operation. Moreover, the various government
bureaus have appeared to contradict themselves on the question of whether the
CCP seeks a specific outcome. The
Department of Homeland Security’s 2025 Homeland Threat Assessment, published
Oct. 2, anticipated that foreign use of “subversive tactics in an
effort to stroke discord and undermine confidence in U.S. domestic
institutions” would increase. Recently,
officials from the ODNI have delivered statements to the press to assert that
Russian cyber actors are attempting to elect Trump and undermine Harris. Yet the
ODNI’s most recent election security fact sheet claims that China “probably does not plan to
influence the outcome” of the U.S. election. Mills thinks
differently and believes that the CCP is “trying to influence the election” to
ensure the election of a candidate who would be less effective at countering
its quest for global hegemony. “What is the
Chinese agenda? I think, as opposed to the Russians, who just want to create
hate and discontent writ large ... it is election interference,” he said. One of the
ODNI’s reports last year revealed that the CCP was more willing to interfere in U.S.
elections now than in previous cycles precisely because it did not believe the
Biden administration would retaliate. The report
said that CCP officials gave operatives more freedom to interfere in U.S.
elections because the regime “believed that Beijing was under less scrutiny ...
and because they did not expect the current administration to retaliate as
severely as they feared in 2020.” . A U.S.
intelligence report last year revealed that the CCP is more willing to
interfere in U.S. elections now than in previous cycles. Frederic J. Brown/AFP
via Getty Images ‘No Such
Thing as Guardrails’ Just as
interpretations of China’s motives have remained murky, so have the various
government agencies responsible for defending Americans from such operations,
which have largely failed to provide any official guidance for how everyday
Americans should identify and respond to such content. In April
this year, Cybersecurity and Infrastructure Security Agency (CISA) senior
adviser Cait Conley said that the agency was ready to help stave off the threat
of foreign influence operations, particularly in the 2024 election cycle. “The
elections process is the golden thread of American democracy, which is why our
foreign adversaries deliberately target our elections infrastructure with their
influence operations,” Conley said in a statement. “CISA is
committed to doing its part to ensure these officials—and the American
public—don’t have to fight this battle alone.” When asked
what Americans can do to identify and counter foreign influence operations,
CISA declined to comment and instead referred The Epoch Times to the ODNI. The ODNI did
not return multiple requests for comment on the matter. When asked
what actions the State Department was taking to address foreign influence in
U.S. elections, a department spokesperson said that it was “focused on the
information environment overseas.” The Epoch
Times has also requested comment from the Department of Homeland Security. Graphika,
whose report does not offer any suggestions on identifying or countering the
content examined, declined to comment. Recorded
Future also did not return a request for comment. In a report published in September, however, the company
suggested that the response to deepfakes should be left to the entities
concerned about reputational damage, whom it encouraged to cooperate with
“fact-checkers, social media platforms, and media outlets.” That’s a
real problem, given the increasing reach of foreign influence campaigns, which,
according to the Recorded Future report, frequently aim to mislead the public
and engage in electioneering. Similarly,
according to research cited in the same report, most people cannot detect
deepfakes and would benefit from guidance on the topic. Indeed,
according to research published in the Journal of Cybersecurity
Education, Research and Practice, most people are simply unable to identify
deepfake videos of people with whom they are unfamiliar, and nearly 30 percent
of people are unable to distinguish deepfakes of people they are familiar with.
Moreover,
even if a person has identified a deepfake video for what it is, they may still
be influenced by it, particularly if it promotes an extreme belief or action. Research published in the academic journal Computers in Human
Behavior found that “false information can have an effect on people’s
political beliefs, even after retraction.” “Even when
people are aware that certain information may not be true, it still impacts
their beliefs and actions,” the report reads. “In other
words, even implausible disinformation can influence political beliefs,
partially beyond the awareness of recipients.” The
prevalence of deepfakes in foreign influence operations could, therefore,
engender a long-term dislike or distrust among American voters for candidates,
even after those Americans find out the information was not real. When asked
what advice he would give Americans, Mills said: “You should be very, very
suspect of anything you see online. “There’s no
such thing as guardrails with what China is up to.”
|