TAIPEI (Taiwan News) — Interior Minister Liu Shyh-fang (劉世芳) on Wednesday defended the government’s investigation into a deepfake video created by former KMT Youth Department Director Wei Chun-yu (韋淳祐), arguing that AI-generated impersonation is not protected by freedom of speech.
Wei created the video at the request of Taipei Mayor Chiang Wan-an (蔣萬安) to promote the KMT’s protest over the government’s handling of the tainted oil scandal, per Liberty Times. The interior ministry said Wednesday that an analysis by the Criminal Investigation Bureau found the video’s narration, which resembles the voice of President Lai Ching-te (賴清德), bore characteristics of AI-generated audio, per UDN.
Liu defended the investigation, saying that while Taiwan protects freedom of speech, impersonating another person’s voice is unacceptable, per Tai Sounds. She said the government has a duty to intervene when deepfakes are used to spread false information, disrupt public order, influence public judgment, or facilitate identity fraud.
Wei said Saturday that police officers visited his home to ask whether he had created the video and whether he was willing to remove it from the platform, per China Times. He argued that anyone seeking to have the video removed should have contacted him directly instead of sending police officers.
National Police Agency Director-General Chang Jung-hsing (張榮興) said the visit was solely to verify the facts and that officers neither asked Wei to remove the video nor sought to restrict his freedom of expression. The interior ministry said Wei’s video may constitute forgery, defamation, and violations of the Personal Data Protection Act, per Liberty Times.
Political reactions largely split along party lines. Presidential Office Spokesperson Karen Kuo (郭雅慧) said that while freedom of speech protects parody, it does not permit the unauthorized use of another person’s voice, per Tai Sounds. Kuo argued that using Lai’s voice in the deepfake was a deliberate attempt to mislead the public.
DPP Taipei mayoral candidate Puma Shen (沈伯洋) said deepfaking the voice of a head of state is typically associated with hostile state actors, per Liberty Times. He said he suspected “there might be people behind the scenes giving the KMT pointers,” adding that the apparent goal was to influence Taiwan’s political landscape.
The KMT, meanwhile, argued that the investigation amounted to suppression of free speech. Chiang said police officers visiting Wei’s home constituted state intimidation and could have a chilling effect on free expression. The KMT legislative caucus echoed that criticism, saying authorities had acted more swiftly in this case than in investigating the tainted oil scandal, per UDN.
TPP Chair Huang Kuo-chang (黃國昌) also criticized the investigation, saying it was intended to deter further parody of Lai, per Liberty Times. He argued that former President Ma Ying-jeou’s (馬英九) administration did not respond in the same way to satirical content targeting Ma.
National Yang Ming Chiao Tung University law professor Chin Mong-hwa (金孟華) called for legislation requiring political AI-generated content to disclose that it was created using artificial intelligence, per SETN. He warned that while parody is protected speech, the spread of deepfake content could undermine democratic decision-making, which depends on informed public debate and accurate information.
