摘要
人工智能研发主体在数据挖掘与机器学习的过程中需要大量地使用作品,其所面临的著作权侵权风险与现实的作品使用需求之间的矛盾,亟待化解。人工智能研发主体对于作品的使用非常广泛,不限于某一特定作品类型,其作品使用行为所可能侵犯的著作权具体权项主要有复制权、演绎权与信息网络传播权。与法定许可相比,将数据挖掘与机器学习的情形纳入合理使用制度之下更具合理性。鉴于当前著作权相关立法在数据挖掘与机器学习的问题上尚处于缺位状态,为应对实践的需要,可以考虑从法律解释的角度入手,尽量将部分合理情形解释入现行合理使用的立法文本之内。具体来说,对于非营利性主体在数据挖掘与机器学习中的作品使用行为,可以将其解释入“个人学习研究”及“学校课堂教学或者科学研究”合理使用之中。后者需要对“少量”作宽松的解释,以特定行业或领域的惯例及研发需求作为判断是否“少量”的实质性标准。对于营利性主体在数据挖掘与机器学习中的作品使用行为,作为应急之策,可以考虑对“科研”作扩大解释,但存在一定的弊端。
Artificial intelligence research entities need to use a lot of works for data mining and machine learning.They therefore face the contradiction between the risk of infringement upon copy-rights and their actual need of using works,which calls for an urgent resolution.The works used by AI research entities are very extensive,not limited to a specific type of work,and the specific copy-rights that may be infringed upon by them mainly include the right of reproduction,the right of de-duction and the right of information network communication.Compared with statutory permits,it is more reasonable to incorporate data mining and machine learning into the fair use system.In view of the lack of copyright-related legislation on data mining and machine learning currently,to meet the needs of practice,we may consider taking the approach of legal interpretation to try to interpret some reasonable situations into the scope of fair use in legislation.Specifically,the behavior of using works by non-profit entities could be interpreted into the fair use of“personal learning and research”and“classroom teaching or scientific research in schools”.The latter requires a loose in-terpretation of“small amount”,which means to take the practice and research requirements of a spe-cific industry or field as the substantive criteria for judging whether the concerned use constitutes“small amount”or not.For the use of works by profit entities,as an emergency plan,we may con-sider expanding the interpretation of“scientific research”.However,this plan has certain draw-backs.
作者
万勇
李亚兰
Wan Yong;Li Yalan
出处
《数字法治》
2023年第3期83-92,共10页
DIGITAL LAW
基金
北京市社会科学基金重点项目(20FXA005)的阶段性成果。