Kneschke vs. LAION: Opening Fresh Perspectives on AI Training and TDM Exceptions

The recent decision in Kneschke vs. LAION by the German Hamburg Regional Court marks a pivotal moment in defining how the CDSM Directive’s text and data mining (TDM) exceptions apply to AI training. In this post, author Martyna Manteghi examines this landmark case, in which photographer Kneschke brought legal action against LAION, alleging copyright infringement for using one of his original works in AI training datasets without authorization. LAION, a nonprofit organization dedicated to advancing AI research by providing open datasets, defended its actions, claiming they were protected under the TDM exceptions outlined in the German Copyright Act and the EU CDSM Directive.

The defendant had obtained the photo from a stock photo website,  freely available to the public, to confirm whether the description offered by a third-party provider accurately reflected the photo’s contents. The final training datasets included only the hyperlinks to the images’ original sources, along with their descriptions and metadata. The Court dismissed Kneschke’s claim of unauthorized reproduction, recognizing that LAION could rely on the TDM exception for scientific research (Art. 3 of the CDSM Directive and Sec. 60d of the German Copyright Act).

Although the decision primarily addressed issues related to the preparatory phase of AI training—such as downloading images—and not the training itself, the Court’s findings provide valuable insights into the practical application of TDM exceptions under the CDSM Directive. This is particularly significant during a time of rapid AI development, where balancing powers in the digital ecosystem is critical to supporting the EU AI sector…