Legal uses of Generative AI

Generative AI has been all the rage in the last year. Its risks of copyright infringement, of bias and discrimination, of fake representations and of safety and trust hazards have been extensively discussed with some anxiety. There is no easy solution at the horizon, but it might be time to start thinking of how Generative AI and its operators could provide some tools to address the issues it creates… and to critically evaluate their promises or commitments for safer and fairer AI applications.

Speakers

• Niva Elkin-Koren, Professor, Tel Aviv University, Faculty of Law

Copyright Regenerated: Leveraging GenAI to define the scope of copyright protection
Generative AI might not only create issues for copyright but provide solutions too. Harnessing GenAI to identify non-copyrighted elements within copyrighted materials could indeed help in properly delineating the scope of copyright protection.

• Sergio Branco, Director of Institute for Technology and Society of Rio de Janeiro

Personal AI systems and a proposal for remuneration of artists 
What compensation strategies for creators in the context of generative AI could be imagined and how contributions to training AI systems can be fairly compensated? Various approaches to remunerating creators for their role in enhancing AI as a tool for creative expression will be explored, with a focus on achieving equitable solutions.

• Kate Klonick, Associate Professor St. John's University Law School

Trust, safety issues and internal policy in OpenAI and Facebook
Based on 40+ hours of interviews with trust and safety, engineers and leadership at OpenAI carried out in the last six months, the red-teaming operations of AI operators and how they address trust and safety issues will be discussed.

Moderated by Séverine Dusollier, Professor, Sciences Po Law School.