Artificial Intelligence (AI) has developed in such a way that it is capable of producing creations and inventions without human intervention through the training of a number of datasets. This normative juridical research aims to look at AI problems from the perspective of AI as a subject and AI results as an object of copyright and patent protection, as well as examining the implications of using creations in datasets to train AI. This research found that AI cannot become a creator and inventor because moral and human rights are reserved for humans, besides that AI cannot take advantage of the economic rights obtained from the protection of creation or patents. This study also found that the use of datasets containing other people’s creations as AI development material has the potential to cause copyright violations. This potential is mitigated by several countries by implementing regulations related to TDM or data scraping for AI machine learning. Finally, this study also found that creations and inventions resulting from AI in general cannot become objects protected by the copyright regime unless they receive direct human contribution or are formulated in statutory regulations such as in the CGW copyright regime in the UK. This research suggests that practices in other countries in copyright and patent protection regimes related to AI can be used as a reference for legal politics in Indonesia to create AI regulations that balance the moral and economic rights of Creators and Inventors with the pace of AI innovation.
Copyrights © 2023