Taiwan deepfake voice row: AI-generated audio targets city leaders in election campaign
Taiwan's Taipei Mayor Chiang Wan-an's campaign team created AI-synthesized deepfake audio of the President's voice for a promotional video, while Tainan Mayor Huang Wei-zhe publicly demonstrated deepfake audio mimicking Chiang's voice to illustrate the dangers of the technology. The Taipei District Prosecutors Office opened an investigation into the original deepfake, with political debate over whether such demonstrations constitute illegal encouragement or legitimate public awareness.
What happened
Taipei Mayor Chiang Wan-an's campaign team commissioned staff to produce AI-synthesized deepfake audio of President Tsai Ching-te's voice for a promotional video related to a July 2026 rally, according to Tainan Mayor Huang Wei-zhe.
Huang Wei-zhe publicly demonstrated AI-generated deepfake audio mimicking Chiang Wan-an's voice at a press conference, creating three synthetic clips purporting to contain statements about Chiang Kai-shek, the 228 Incident, and White Terror, alongside one authentic recording of Chiang Wan-an's real voice.
Chiang Wan-an stated the deepfake incident involved his campaign's junior staff and told critics to hold him accountable directly, saying 'if there's an issue, come at me, not my staff.'
Reported impact
Taipei District Prosecutors Office received a report from the Criminal Investigation Bureau alleging the video suspected use of AI to synthesize President Tsai's voice and potentially violated forgery laws, opening an investigation.
The incident sparked political debate during the 2026 Taipei mayoral election campaign, with Kaohsiung Mayor Chen Chi-mai condemning both the original deepfake of the President and Huang's demonstration as 'very inappropriate' and calling for prosecutors to investigate quickly.
Index analysis
Huang Wei-zhe stated his purpose was to demonstrate the dangers of deepfake technology and raise public awareness, comparing it to government anti-fraud awareness campaigns that demonstrate scam tactics.
Taipei mayoral candidate Shen Po-yang distinguished between warning the public about deepfake risks and encouraging deepfake creation, arguing that endorsing deepfake use of presidential or senior official voices crossed a legal line.
Governance controls
Chiang Wan-an characterized the prosecution's investigation as 'state violence' and alleged it was motivated by political targeting from the Democratic Progressive Party, arguing the case should not become a tool for political persecution.
Classification
Sources and evidence
This record was researched and written by the Index, based on the following original reporting.