AI 스피커 음성정보의 합리적 규제 방안에 대한 연구

AI speakers can create various social benefits in the aspects of usefulness. However, current regulation on voice data is too comprehensive to consider identifiability differences among biometrics and different level of risk of privacy breach and it possibly causes reverse discrimination between domestic and foreign business operators.
Since voice data is subject to the same regulations with other types of biometrics such as fingerprint and iris although it has lower identifiability and less risk of privacy breach, it is limited to promote AI speaker industry and enhance user benefits. In particular, as the rule of disposal of original data and the rule of additional consent specified in the current regulation on voice data can cause conflicts among the benefit and protection of other laws and regulations, a reasonable regulation approach is needed to ensure the benefit and protection of the law.
As a reasonable regulation approach, the right to opt-out and self-regulation are suggested in this article. In case of voice data that has lower personal identifiability and less risk of privacy breach, it is necessary to deregulate by guaranteeing the right to opt-out rather than abrogating the rule of disposal of original data and the rule of additional consent. Furthermore, business operators and users can build mutual trust and reverse discrimination between domestic and foreign business operators can be solved if we at first apply self-regulation, a type of soft law flexible towards the development of technology, rather than hard law.