Time: 12:00-1:00 PM ET
Date: Friday, September 18, 2026
Open-weight AI models – which the public can download and adapt freely – are rapidly closing the gap with proprietary systems like ChatGPT, Claude, and Gemini. They offer real benefits: lower cost, control over data, and greater customization. These systems are and will continue to be a critical component of AI adoption around the world.
However, there are real issues: too much of the investment in the open ecosystem has been directed at Chinese models. And open release also means the capabilities circulate permanently – dual use concerns are real and current safeguards can be stripped out with relative ease.
Unfortunately, the debate about open-weight models typically devolves quickly. Conversation reduces to either “lock it down,” which ignores the reality of the market and geopolitics, or “don’t do anything,” which neglects capability concerns while foreclosing public investment.
This panel will explore potential paths to restoring U.S. leadership in the open model ecosystem while mitigating these pitfalls. Panelists will discuss emerging efforts in the public and private sector, and how we can ensure that Americans are ready to capture the maximum benefit from this type of AI.
Panelists:
- Ian Reynolds, AI Public Policy Manager, Hugging Face
- Joseph Brandifino, Global Head of Policy, Reflection
- Miranda Bogen, Chief Technologist & Director, AI Governance Lab, Center for Democracy & Technology
- Josh New, Director of Policy, SeedAI (Moderator)