Chiang Wan-an defends 'deepfake' as free speech; lawyer warns of criminal charges
Translated from Chinese, summarized and contextualized by DistantNews.
At a glance
- Taipei Mayor Chiang Wan-an defended the use of AI-generated deepfake audio mimicking President Lai Ching-te as freedom of speech.
- A lawyer argued that creating and distributing such content could lead to criminal charges for document forgery and violations of personal data and food safety laws.
- The lawyer stated that while foreign examples exist, Taiwan's legal interpretation of "parody" is strict, and intent to harm can override free speech protections.
Taipei Mayor Chiang Wan-an has characterized the police investigation into the creation of a "deepfake" video mimicking President Lai Ching-te's voice as "state violence." The video, produced by former KMT new media director Wei Chun-yu, used AI to simulate Lai's voice for a clip titled "You Can't Oil Me." Chiang invoked examples of AI-generated content in U.S. and British media, asserting that such mimicry falls under freedom of speech.
However, lawyer Fang Yen-hui presented a contrasting legal perspective, suggesting that the act could result in both criminal and civil penalties. Fang explained that creating and distributing AI-generated audio and video mimicking a person's voice constitutes forgery of a private document under Taiwan's Criminal Code. This could lead to imprisonment of up to five years.
Furthermore, Fang noted that a person's voice is considered personal data under the Personal Data Protection Act. Unauthorized collection or processing of such data, without a specific purpose or consent, could also result in criminal charges and fines. The content of the video, which touched upon food safety issues, could also trigger violations of the Food Safety and Health Management Act, carrying penalties of up to three years imprisonment or fines.
Regarding Chiang's defense based on freedom of speech and foreign examples of AI satire, Fang cautioned that Taiwan's legal system takes a strict approach to "parody" or "satire." He emphasized that if the creator knowingly disseminates false information with the intent to harm another person's interests, the act transcends freedom of speech, even if labeled as AI-generated. The key factor, Fang stated, is whether the content causes the public to be "misled."
You Can't Oil Me
Originally published by Liberty Times in Chinese. Translated, summarized, and contextualized by our editorial team with added local perspective. Read our editorial standards.