Taiwan National Police Agency reports on handling of election disinformation and deepfake forensics
Taiwan's National Police Agency reported progress on investigating and handling election disinformation and deepfake audio and video. As of now, police agencies nationwide have investigated and handled 219 related cases, of which 134 cases involving 176 people have been referred to judicial authorities for investigation. Police are strengthening their technological investigative capacity, using AI deepfake detection, digital forensics and online intelligence analysis to quickly verify suspected illegal content and trace its source. For audio, video and image posts of unknown origin, with sensational content or suspected to be AI-generated, the public should first cross-check through official and trusted channels and should not share them before verification; candidates who find deepfake content suspected of impersonating their own voice or image may apply to the Criminal Investigation Bureau for forensic identification, and after confirmation may submit a request for handling through the Ministry of Digital Affairs' "Election AI Deepfake Information Reporting Network"; platforms receiving such requests should restrict viewing, remove or take down the content within 2 days. Under Article 104 of the Civil Servants Election and Recall Act, spreading false information with the intent to cause a candidate to be elected or not elected, sufficient to harm the public or others, is punishable by up to 5 years' imprisonment; committing the offense by spreading or broadcasting a candidate's deepfake audio or video is punishable by up to 7 years' imprisonment.