Cybersecurity teams have uncovered a vast network of AI-driven social media accounts linked to China, raising alarms about possible foreign interference as the 2024 elections approach.
The cybersecurity firm CyberCX revealed that nearly 8,000 fake X accounts utilized generative AI to spread politically divisive content ahead of the upcoming November vote.
This network, dubbed “Green Cicada,” employed AI to transform pre-existing political posts into seemingly authentic retweets and replies, thereby enhancing the visibility of already divisive messages on the platform.
Matt Cardy/Getty Images
Operating since at least April, CyberCX classifies Green Cicada as one of the largest known networks of inauthentic activity. The activity has surged since July, as reported by National Security News.
Most accounts within this network remain inactive, being dormant thirty times more than the active portion. Experts warn that this “sleeping giant” could wake up as the elections draw nearer.
The network’s disinformation campaigns show no clear political allegiance; they’ve targeted figures from various sides, including two separate campaigns in August that criticized and supported Kamala Harris around her nomination as the Democratic candidate.
GDA via AP Images
Although CyberCX believes the current state of Green Cicada is unlikely to sway public opinion, they highlighted that future, more sophisticated AI campaigns could produce harmful content efficiently and with little human oversight. Detecting such campaigns might become extremely challenging for platforms like X.
The report states the campaigns also amplify contentious political discussions in the UK and other democracies, showcasing clear ties to China.
According to CyberCX, “the controlling system behind the network is closely linked to China,” including associations with a Chinese-language AI model and connections to Tsinghua University and Zhipu AI, a well-known Chinese tech company.
While direct links to the Chinese government remain unverified, the strategy aligns with long-standing Chinese tactics for information warfare.
This isn’t the first instance; in 2023, Meta dismantled around 8,000 Facebook accounts tied to a Chinese spam operation that forwarded pro-China sentiments and criticized US policies.
Andrew Harnik/Getty Images
Concerns over foreign meddling have already surfaced in this election cycle. Microsoft’s Threat Analysis Center released a report in August detailing the “foreign malign influence” from China, Russia, and Iran as the elections loom.
Many of these operations were aimed at influencing US voters, raising significant concerns about the integrity of the upcoming elections.
Recently, the U.S. Department of Justice charged two Russian employees from the state-controlled news agency RT with violating the Foreign Agents Registration Act as part of a $10 million scheme to infiltrate U.S. media with covert propaganda.
Attorney General Merrick Garland emphasized, “The Justice Department will not tolerate attempts by authoritarian regimes to exploit our country’s free exchange of ideas to further their propaganda agendas.”