Webinar: Copyright and Large Language Models

Get the latest updates on the training and use of large language models (LLMs) from the perspective of law, computer science and application. Dr. Mehdi Ali, senior scientist at Fraunhofer IAIS and the Lamarr Institute, will provide insights into the training of LLMs using the example of Teuken-7B, a large European AI language model developed as part of the OpenGPT-X research project. Dr. Maximilian Poretschkin, Head of the Innovation Group “Trustworthy AI” at the Lamarr Institute, as well as other renowned experts from science and practice will address legal issues in their presentations, in particular possible copyright infringements in the output of LLMs.

The topics at a glance:

• Copyright infringements in the output of LLMs: a quantitative benchmark
  Prof. Dr. Anna Bernzen, University of Regensburg | Dr. Maximilian Poretschkin, Fraunhofer IAIS
• Insights into the training of LLMs using the example of TEUKEN
  Dr. Mehdi Ali, Fraunhofer IAIS & Lamarr Institute
• Infringement in the output of LLMs
  Dr. Kai Welp, legal advisor GEMA

The webinar will be conducted in German language and is organized by the project “Zertifizierte KI” (Certified AI) and supported by the Lamarr Institute as well as KI.NRW, the University of Regensburg, Fraunhofer IAIS and the University of Bonn.