AI systems are playing an increasing role in society, but have also given rise to several practical problems, such as producing output (e.g. realistic images, artworks, poems) or causing damages. Accordingly, AI systems are challenging established legal doctrine and raising policy questions. This research poses the central research question on who owns AI and AI-output. First, the question will be considered whether AI systems can be legally qualified as goods in the sense of property law, which is challenged by AI systems’ complex and immaterial nature. At the same time, AI systems might be protected by intellectual property rights, namely copyright and patent law, while the relationship between these two classifications is not clear. Second, since AI systems are engaging questions on (non-)contractual liability issues, it will be discussed whether legal personhood should be attributed to AI. Third, it will be examined whether copyright and patent law should apply to output generated autonomously by AI systems. The innovative character of the project lies in combining different legal branches by examining interrelated questions together. The research methods will include critical analyses of legislation, case law and legal literature, combined with comparative law methods when useful. Due to the economic value of both AI systems and AI-output, it is essential to establish the applicable law.