Microsoft has issued a warning regarding the potential use of artificial intelligence by Chinese state-backed cyber groups to disrupt major elections in the United States, South Korea, and India. This forecast follows a “test run” during Taiwan’s presidential election, where AI-generated content was utilized in an attempt to sway voters away from a candidate favoring sovereignty.
The tech giant’s threat intelligence team published findings suggesting that in addition to Chinese cyber actors, North Korean groups might also engage in similar activities targeting these high-profile elections. The primary method expected to be used involves the creation and dissemination of AI-generated content across social media platforms, aimed at supporting China’s political interests in these countries.
Microsoft’s analysis highlighted that China’s experimentation with AI in the political arena is not new. During Taiwan’s recent presidential election, a Beijing-backed group, identified as Storm 1376 and also known by names such as Spamouflage or Dragonbridge, was particularly active. This group attempted to influence the election outcome by distributing fake audio clips and AI-generated memes critical of William Lai, the pro-sovereignty candidate who ultimately won. These efforts included unfounded accusations against Lai and utilized AI-generated TV news anchors to spread disinformation.
One specific instance cited by Microsoft involved a fake audio clip on YouTube, purportedly of election candidate Terry Gou endorsing another candidate, which was likely generated using AI technologies. Although YouTube removed the content promptly, this instance marks a significant moment where AI-made content was deployed in an effort to influence a foreign election.
Moreover, Microsoft’s report touches on ongoing Chinese influence campaigns within the United States, where social media accounts backed by Beijing have been observed posing provocative questions aimed at identifying divisive issues among US voters. This strategy could be part of a larger effort to fine-tune China’s disinformation campaigns ahead of the US Presidential election.
Despite the current assessment that AI-generated content has a low impact on swaying public opinion, Microsoft warns of the potential effectiveness of these tactics in the future as China continues to refine its methods.
Expanded Coverage: