Podcast series “KPMG Law on air”: Generative AI

Generative AI only entered the world of work a few months ago and it is already clear that it will significantly change many job profiles.

Technical progress is much faster than its regulation. EU legislators have also been overtaken by technological developments. For more than two years, the EU Commission had been working on an AI Act and then had to add regulations on generative AI to it. The fact that there is hardly any case law on AI so far also leads to legal uncertainty. Above all, copyrights and data protection regulations could be violated with Large Language Models.

An AI tool is a black box in terms of data protection law

In terms of data protection law, it is problematic that AI tools are like a black box. Users usually do not recognize which data is processed and which algorithms are used. This poses enormous risks. A company can only be compliant if there is a certain degree of transparency. Employees who work with personal or even sensitive data must pay particular attention.

Dr. Anna-Kristine Wipper and Francois Heynike talk in the podcast about the legal classification of generative AI.

The core topics of the podcast:

• From minute 0:52: Why the topic of AI has become much more important
• From minute 2:45: The AI Act and the integration of generative AI.
• From minute 3:44: Copyright pitfalls in the use of generative AI.
• From minute 5:44: Copyright of images created with AI
• From minute 8:48: Privacy risks of generative AI.
• From minute 12:52: Legislative requirements

All “KPMG Law on air” episodes can be found here.