Abstract
This paper examines the issue of deepfake technology from technological and ethical perspectives. This technology has the potential to revise the traditional notion of replication and undermine the evidentiary capabilities of video and audio. The development of deepfakes also risks creating a threat to authentication technology. Pornographic videos using deepfake technology have also become a social problem. Therefore, this paper examines what possible countermeasures can be taken to address the risks of this new technology. Moreover, we argue that engineers who develop AI programs should consider the risks seriously.
