The State of Frontier AI Evaluation and Red Teaming
ABOUT THIS EVENT
- WHO IT IS FOR: AI researchers, safety engineers, policy makers, and industry professionals interested in model security and robust evaluation.
- WHAT IT IS ABOUT: An in-depth exploration of the current landscape, limitations, and future challenges of frontier AI evaluation and red teaming.
- FORMAT: A technical opening talk followed by a moderated multi-disciplinary panel and a networking session.
Join Gradient Gatherings, co-hosted by the Gradient Institute and FAR.AI, for an essential deep dive into the evolving world of frontier AI safety. Held during the NeurIPS 2026 week in Sydney, this event examines the cutting edge of model evaluation and red teaming to uncover what current methodologies reveal and where the critical gaps remain.
The evening features a technical opening presentation designed to set the stage for a high-level moderated panel. Bringing together diverse perspectives from research, government, and industry, the discussion will tackle the most pressing challenges in ensuring AI robustness. Stay after the session to connect with fellow experts and practitioners in the field during our dedicated networking time.
EXPLORE MORE