This research compared penal laws related to deepfake technology across various jurisdictions to identify optimal legal solutions for managing AI-facilitated crimes. 

 Some jurisdictions, such as South Korea and Virginia in the United States, penalize creating and distributing sexual deepfakes. However, the appropriateness of the penalties, given the grave harm to the victims, is still debated. Other jurisdictions, like Texas, penalize using deepfake technology to spread false information about public elections or politics. This has gained particular significance because deepfake videos, realistically portraying events that never occurred, can easily allow people to believe and spread false information. Analyzing penal laws related to deepfake technology reveals the challenges of developing legal frameworks that can effectively address the rapidly evolving landscape of AI-facilitated crimes. 

 Based on the review of the laws, three legislative recommendations are made, including (1) a periodic review of the development of socially undesirable use of AI, (2) an examination of legal frameworks that govern conventional offenses, such as sexual offenses and fraud, to ensure that they can flexibly address new types of AI-facilitated crimes, and (3) a reassessment of the appropriateness of penalties for existing laws on AI crimes.