Exploring permissible data use for AI model training 

Morrison Foerster’s Lokke Moerel and Marijn Storm break down the European Data Protection Board’s ChatGPT Taskforce interim report. They discuss the implications of the ruling, in particular what it means if large language models cannot be trained on special sets of personal information containing ethnic origins, political opinions, sexual orientations and more.
Full story

​ 

Leave a Reply

Your email address will not be published. Required fields are marked *