Whether or not the training of AI models with copyright-protected data is legally permissible is a controversial issue and the subject of ongoing debate.
In its decision of 11th November 2025, the Regional Court of Munich I (case no. 42 O 14139/24) commented on this issue and decided that training AI models with copyright-protected works constitutes exploitation of copyright-protected works, for which a licence is required. Further, it held that the use of copyright-protected content in the output of an AI also generally constitutes a relevant act of exploitation, meaning that a license is required for this as well.
Magnus Brau
Email: magnus.brau@rittershaus.net
Dr. Markus Spitz
Email: markus.spitz@rittershaus.net