China’s AI Influence Campaigns: A Threat to East Asia and the United States?

Microsoft has recently released a report stating that China is using AI-generated media to influence elections in East Asian countries and the United States. The report, issued by Microsoft Threat Intelligence, highlights several instances where Chinese propaganda actors created memes, videos, and audio using artificial intelligence to spark outrage and impact elections in East Asia. Additionally, Chinese government-linked accounts have been spreading AI-generated content with the intention of inciting chaos in the US.

One AI-generated image that the report incorrectly attributes the 2023 fires in Hawaii to a US-produced “weather weapon.” Another piece of media aimed to create distrust in the government by linking a train derailment in Kentucky in November to sinister motives. Many social media messages originating from Chinese accounts ask followers to disclose their preferred presidential candidates, suggesting an effort to gather intelligence on key voting demographics in the US.

While the threat level is currently considered low, the report warns that China’s AI-driven influence campaigns could become more effective in the future. This report comes after the US Office of the Director of National Intelligence previously released an assessment detailing China’s efforts to undermine democracy, sow doubt in US leadership, and expand Beijing’s influence.

Both Microsoft’s report and the government’s assessment highlight the existence of “Spamouflage” or social media accounts linked to the Chinese Communist Party propagating divisive, often AI-generated content. The analogous tactics used by China in recent influence campaigns are likened to those employed by Russia in the 2016 elections, where troll farms disseminated election disinformation.

In a separate report by Reuters, it was revealed that former president Trump had authorized a CIA program aimed at discrediting the Chinese government on social media during his tenure in office.