TAIPEI (Taiwan News) — AI is enabling a new level of interference in Taiwan’s elections, the DPP warned on Sunday.
The party’s China affairs department said in a Facebook post that AI-generated content already emerged in the 2024 general election, including fake videos of the “secret history” of Tsai Ing-wen (蔡英文) and the spread of AI-generated images and memes across social media platforms.
As AI technology advances, the risks of interference will increase, the party said.
The DPP highlighted three key areas of concern regarding AI-enabled interference, including precision targeting, identity impersonation, and last-minute attacks.
Precision targeting entails analyzing personal data, social media interactions, and online behavior to tailor political messages that are most likely to persuade specific demographic groups based on age, occupation, location, and political leanings, the party said.
Identity impersonation uses AI to mimic the language and tone used by Taiwanese to operate fake accounts or virtual personas, infiltrate online communities, and make foreign operators appear as domestic voices.
Last-minute attacks refer to releasing deepfake audio, videos, false endorsements, or fabricated scandals in the days or hours before voting. This is intended to exploit limited time for verification to influence voter decisions.
The DPP urged the public to remain vigilant, especially when encountering sudden political revelations, social media content, or major candidate statements before elections.
The party also called for stronger deepfake detection, AI content labeling and traceability, election cybersecurity, and other preventative measures. In particular, it stressed the need to shorten the time needed to detect, verify, and clarify fake social media content.
The warning comes as the Chinese Communist Party has reportedly designated Taiwan’s 2026 local elections as a key focus of its coercive efforts toward Taiwan, per the DPP.
