Taiwan's ruling Democratic Progressive Party (DPP) has warned that China is comprehensively upgrading its methods of interfering in Taiwan's elections through artificial intelligence — bad news for the security of Taiwan's elections and for its democracy.

• The DPP says the CCP has made this year's year-end nine-in-one local elections a priority in its work targeting Taiwan, raising the risk of outside interference.
• AI makes disinformation harder to tell apart: it is used to run fake accounts that mimic the tone of Taiwanese people and to infiltrate online communities, making foreign operations look like the voices of Taiwan's own public.
• The most dangerous is the "final surprise attack" — releasing deepfake audio of candidates, fake videos, or fabricated scandals in the hours before voting, leaving no time to verify or set the record straight.
• The DPP describes this as having formed a complete "election-interference industry chain": the front end gathers public-opinion and voter data, the middle deploys messages through local collaborators, and the back end analyzes the effects and adjusts the strategy.
• It advises the public to remember the "three don'ts and one check": don't trust content whose source is unknown, that deliberately stirs up emotions, or whose original version cannot be found — and verify the original source first.

Outlook: With the year-end elections approaching, this kind of AI-driven disinformation offensive is expected to intensify on the eve of voting, and the pressure on government and civil-society fact-checkers will keep rising.