Experts from academia, the tech industry, and civil society will discuss the type of speech AI should be permitted to generate, and how to balance freedom of expression and safety. Crucial to this discussion is the debate surrounding open- and closed-source models in Generative AI. Some experts have argued that open-source models provide greater freedom to generate content, and can combat market concentration, spur innovation, and enhance transparency. However, others respond that open-source grants uncontrollable power to the public at large, including bad-faith actors, and that robust and significant guardrails are needed. This panel will discuss different approaches to balancing free speech and safety, the risks and opportunities that open-source models bring versus closed-source models, and the lessons we can learn from other fields, such as social media.


Jordi Calvet-Bademunt, Senior Research Fellow, The Future of Free Speech at Vanderbilt University (Moderator)

David Harris, Chancellor’s Public Scholar, University of California, Berkeley

Sarah Shirazyan, Content Policy Manager, Meta Inc. and Lecturer in Law, Stanford University

Peter Cihon, Senior Policy Manager, GitHub



July 23, 2024


11:10 AM — 12:00 PM PT


San Fransisco, CA

Learn More

Add to Calendar