Regulating Large Language Models - A Roundtable Report - Paul Friedl Gabriel Nicholas

The NYU Information Law Institute is an academic center for the study of law, policy, and social norms defining and affecting the flow of information in a digitally networked society. [...] The day concluded with a Co-hosted by the NYU brainstorming session asking participants to debate the potentials and downsides of different regulatory School of Law Information interventions that could be used to address the Law Institute and the problems raised earlier in the day. [...] They argued that information regarding the sources of training data could enable researchers to have a better understanding of the Any transparency potential shortcomings of a specific system, and could obligations would have also reduce barriers for bringing (legitimate) legal to be tailored to their actions. [...] Such mechanisms could include updated robot exclusion protocols25 or new licensing and compensation structures.26 24 In the EU, on the other side, Articles 3 and 4 of the Copyright in the Digital Single Market Directive protect Text and Data Mining. [...] However, in addition to the topics and regulatory proposals discussed above, the roundtable also yielded a number of important general takeaways on how to approach and frame regulatory conversations about large language models.