Time: 9:00 – 10:30 am ET
Date: Thursday, January 29, 2026
Current evaluation approaches used by AI model developers often fall short when tasked with assessing and improving model performance in non-English languages and contexts. Researchers around the world are developing community-led benchmarks and evaluation tools which prioritize participation by native language speakers and context experts to more effectively evaluate the usefulness and safety of major AI models, yet many of these efforts are happening in silos.
In the lead up to the India AI Impact Summit in February 2026, the Center for Democracy & Technology, Cornell Global AI Initiative, and Masakhane’s African Languages Hub will host an officially recognized pre event to the Summit to bring together researchers, policy experts, model developers, and other stakeholders to exchange insights on making evaluations more multilingual and surface avenues through which this work can occur. Some of the barriers impeding more robust multilingual evaluations — data cost and availability, for example — exist across languages, so exchanging lessons and tactics can help advance evaluation tooling in multiple languages and contexts. The ultimate goal: to make evaluation more representative of non-English languages and contexts and make tools readily available to model developers and deployers.
This official pre-summit convening will begin with virtual presentations by leaders working on advancing multilingual evaluation efforts around the world. After the presentations, a panel discussion will dive into how these efforts should be supported and what gaps exist to making these evaluation tools available to model developers and deployers, ranging from industry to public actors.
Featured Speakers:
- Alexandra Givens, President and CEO, Center for Democracy & Technology
- Shri Abhishek Singh, Ministry of Electronics and Information Technology, Government of India
- Miranda Bogen, Director, AI Governance Lab, Center for Democracy & Technology
- Chenai Chair, Director, Masakhane African Languages Hub
- Bennett Hillenbrand, MLCommons
- Sara Hooker, Co-Founder, Adaptation Labs
- Faisal Lalani, Head of Global Partnerships, Collective Intelligence Project
- Chinasa T. Okolo, Founder, Technecultura
- Roya Pakzad, Fellow, Mozilla Foundation
- Sunayana Sitaram, Principal Researcher, Microsoft Research
- Dhanaraj Thakur, Director, Emerging Technology Initiative-Multiracial Democracy Project Washington University
- Aditya Vashistha, Assistant Professor, Cornell University
Accessibility: If you have questions, concerns, or access needs, please contact events@cdt.org in advance of the event.