Taiwanese mayors clash over AI-generated audio, raising deepfake concerns

Tainan Mayor Huang Wei-che released an AI-generated audio clip mimicking Taipei Mayor Chiang Wan-an's voice, accusing Chiang of deliberately misleading the public. Huang presented three such audio clips, criticizing Chiang, a lawyer himself, for his alleged falsehoods.

In response, Taipei City Government's Research, Development, and Evaluation Commission Chairperson, Yin Wei, accused the ruling Democratic Progressive Party (DPP) of exhibiting double standards. Yin questioned when the Criminal Investigation Bureau would investigate Huang Wei-che.

Lawyer Huang Ti-ying commented on the situation, stating that Chiang Wan-an's previous assertion that deepfakes are legal, provided they are labeled, implies a general consent for the use of his voiceprint. According to Huang Ti-ying, if an individual consents to the use of their voiceprint and electronic records, and the deepfake is properly disclosed, it does not violate personal data protection laws or criminal statutes.

Huang Ti-ying further elaborated on Facebook that Chiang Wan-an's claim of deepfakes being legal, when labeled, constitutes a broad agreement for the use of his personal voice data. He asserted that this consent is specific to Chiang's own voiceprint and cannot be extended to others. Therefore, unless Chiang retracts his stance on the legality of labeled deepfakes, his general authorization for the use of his voice data remains in effect.