In this episode of Legal4Tech - The Podcast, we sit down with Assistant Professor Paulina Jo Pesch to unpack how generative AI really works and what that means for copyright and data protection law.

We explore topics such as:

  • How LLM are trained on text, probabilities and "temperature", beyond the myth of superintelligence.
  • Whether training data, outputs and even the models themselves can count as personal data under EU data protection rules.
  • Recent German court decisions on song lyrics, text and data mining and what they signal for creators, platforms and AI developers.
  • The gap between hype, technical reality and legal safeguards in the age of generative AI.


Join us for a clear deep dive into the legal and technical foundations of generative AI. Shownotes:

  1. Prof. Dr. Paulina Jo Pesch Bio
  2. Gema v. OpenAI
  3. Gema v. Suno

Sponsored by Digibeetle

Digibeetle helps privacy, AI, and data professionals stay on top of EU digital law, with notifications, a daily dashboard, and powerful search across thousands of sources. Learn more at ⁠⁠⁠⁠⁠https://digibeetle.eu⁠⁠⁠⁠⁠.


Sponsored by GIEF.

The Geopolitical Insight and Education Foundation is an independent policy and research think tank focused on stability in our democracies. Learn more at https://www.giefoundation.net/

Rss
Podden och tillhörande omslagsbild på den här sidan tillhör Legal4Tech. Innehållet i podden är skapat av Legal4Tech och inte av, eller tillsammans med, Poddtoppen.