인공지능 시대의 저작물 대량 디지털화에 관한 소고

In copyright system, ex ante permission should be secured to exploit a copyrighted work. Any use of copyrighted work without prior permission shall constitute a copyright infringement, if there is no justification of that use such as a limitation and exception to copyright. When it comes to the mass digitization of works, it is extremely difficult or even impossible to get prior permission to use from copyright holder due to high transaction cost. Thus, as long as the current copyright regime persists, there will be a certain tension between copyright system and mass digitization. Here, this article studies the viable solutions considering court precedents and laws of some countries to solve this matter. This includes temporary copying, fair use doctrine and specific exceptions on Text/Data Mining (TDM). However, mass digitization cannot easily meet the requirements of temporary copying exception and fair use doctrine seems to be unpredictable in terms of application. Therefore it is needed to introduce new copyright exception, i.e, Text/Data Mining (TDM). We need to address existing laws without hesitation, if we can enjoy the massive benefit of mass digitization of works.