Disinformation is on the rise. How does it work?

In January 2024, in the run-up to elections in Taiwan, hundreds of video posts appeared on YouTube, Instagram, X and other social platforms entitled “The Secret History of Tsai Ing-wen”. News anchors, speaking English and Chinese, made a series of false claims about Ms Tsai, the outgoing president, and her ruling party. On election day itself, January 13th, an audio clip began to circulate in which Terry Gou, a candidate who had dropped out of the race in November, seemed to endorse the candidate of the China-friendly KMT party (in fact, Mr Gou made no endorsement).

Both the video clips and audio were probably created using artificial intelligence (AI) and posted by a Chinese state-backed propaganda group known variously as Spamouflage, Dragonbridge and Storm-1376. In a report released on April 5th, the Threat Intelligence team at Microsoft, a tech firm, said this was the first time it had seen a nation-state use AI-generated material to sway a foreign election.