Beijing conducted a trial in Taiwan, using AI-generated content to sway voters from a pro-sovereignty candidate
Microsoft has warned that China will use AI-generated content to disrupt elections in the US, South Korea, and India, following a trial during Taiwan’s presidential poll. The tech firm anticipates Chinese state-backed cyber groups, along with North Korea, targeting high-profile elections in 2024.
Microsoft has warned that China plans to disrupt elections in the US, South Korea, and India using AI-generated content, following a trial during Taiwan’s presidential poll. The tech firm expects Chinese state-backed cyber groups, along with North Korea, to target high-profile elections in 2024.
Microsoft stated that China would, at the very least, produce and disseminate AI-generated content via social media to support their positions in these significant elections. The company noted that while the current impact of AI-generated content is limited, this could change.
“While the current impact of such content in influencing audiences is minimal, China’s increasing experimentation with memes, videos, and audio augmentation will persist—and could become more effective in the future,” Microsoft cautioned.
According to Microsoft’s report, China had already tested an AI-generated disinformation campaign during the Taiwan presidential election in January. This marked the first instance of a state-backed entity using AI-generated content to influence a foreign election, as per the company.
During the Taiwanese election, a Beijing-backed group known as Storm 1376, also called Spamouflage or Dragonbridge, was highly active. They attempted to influence the election by posting fake audio on YouTube of candidate Terry Gou endorsing another candidate, which Microsoft believes was likely AI-generated. YouTube removed the content promptly.
The group also circulated a series of AI-generated memes about the successful candidate, William Lai, making baseless claims of embezzlement. Additionally, they utilized AI-generated TV news anchors to spread unsubstantiated rumors about Lai’s personal life, similar to tactics used by Iran.
Microsoft reported that the AI-generated news anchors were produced using the CapCut tool, developed by the Chinese company ByteDance, which owns TikTok.
Microsoft also noted that Chinese groups are still conducting influence campaigns in the US. They are using social media accounts to pose divisive questions and try to understand the issues dividing US voters.
“This could be an effort to gather intelligence and gain precision on key voting demographics ahead of the US Presidential election,” Microsoft stated in a blog post accompanying the report.
On X, formerly Twitter, one post discussed a $118 billion bipartisan bill in the US that included $20 billion for the US-Mexico border and a $75 billion package for Ukraine and Israel, asking for reactions. Another post highlighted the loss of an F-35 fighter in South Carolina, attributing it to the Biden administration and seeking opinions.
The report coincided with a statement from a White House-appointed review board, which attributed a breach of senior US officials’ email accounts to errors by Microsoft. Last month, the US and UK governments accused China-backed hackers of a prolonged cyber campaign targeting politicians, journalists, businesses, and the UK’s election watchdog.