Back to issue
Cover of: Haftungszuweisung für Urheberrechtsverletzungen durch generative KI-Systeme
Jonathan Pukas

Haftungszuweisung für Urheberrechtsverletzungen durch generative KI-Systeme

Section: Aufsätze
Volume 16 (2024) / Issue 1, pp. 38-69 (32)
Published 07.05.2024
DOI 10.1628/zge-2024-0004
Published in German.
Due to a system change, access problems and other issues may occur. We are working with urgency on a solution. We apologise for any inconvenience.
Summary
When using generative AI, copyright infringements may occur from generating creative output such as text or images. For these copyright infringements, liability must be assigned to one or several of the parties involved. However, the question of how to allocate copyright liability cannot be determined just by analyzing current copyright law. Rather, it is necessary to examine who should bear the liability risks for copyright infringements by generative AI already from a regulatory perspective. This allocation of liability depends on the functions and objectives of the liability regime. During the analysis based on these liability goals, the different legal consequences that may be triggered must also be distinguished - damages on the one hand, as well as injunctive relief and removal on the other. As a result, the liability risks should be allocated based on case groups, and parties may be obliged to jointly bear the liability risks. In the article, the allocation of risk according to liability theory provides the framework for an analysis of current copyright law. In general, no systemic deficits in the copyright system of legal remedies can be found for the allocation of liability for copyright infringements in the output of generative AI.