Japan’s ‘Traitor Takaichi’ hit with online smear campaign using fake accounts
Using AI-generated images and simplified Chinese characters, a network of 3,000 fake accounts reportedly tried to sway Japan’s election
A study by Tokyo-based Japan Nexus Intelligence, which analyses digital public discourse, found that some 3,000 accounts had been actively posting malicious content about Takaichi since late January.
The activity surged roughly a week before campaigning for the House of Representatives election began on January 27, the Yomiuri Shimbun newspaper reported on Monday.
Posts in both English and Japanese circulated a range of claims, including that “the prime minister has opened the path to military expansion and historical revisionism”, that she “bought votes from the Unification Church” and that “social security burdens on the younger generation are increasing”.
The Japanese-language messages exhibited awkward phrasing and linguistic quirks that suggested machine translation, as well as the use of simplified Chinese characters not found in Japan. The account names blended Japanese katakana script with Chinese characters, known as kanji.
Roughly one-third of the accounts generated original posts, while the rest amplified them through reposts. Most had only a few entries each, a pattern analysts said reflected an effort to avoid detection.