Mon 15 - Fri 19 June 2026 Boulder, Colorado, United States

Track

Time Zone

The program is currently displayed in (GMT-06:00) Mountain Time (US & Canada).

Use conference time zone: (GMT-06:00) Mountain Time (US & Canada)Select other time zone

The GMT offsets shown reflect the offsets at the moment of the conference.

Time Band

By setting a time band, the program will dim events that are outside this time window. This is useful for (virtual) conferences with a continuous program (with repeated sessions).
The time band will also limit the events that are included in the personal iCalendar subscription service.

Display full programSpecify a time band

Save

When

Mon 15 Jun 2026 09:10 - 10:10 at Meadows B - Keynote Chair(s): Shraddha Barke

Abstract

Formal Methods for Frontier AI Systems

Formal methods are often dismissed as too rigid, complex, or unscalable for frontier AI systems, including LLMs, VLMs, and agentic systems. In this talk, I will challenge this common assumption through both theoretical insights and empirical evidence across a range of domains, including chatbots, autonomous driving, mathematical reasoning, code generation, and agentic AI.

I will present a new set of efficient formal frameworks for frontier AI systems that:

Certify safety properties such as secure code generation and catastrophic conversational risks, yielding stronger generalization guarantees than standard evaluation methods, including benchmarking and red teaming.
Guide generation with formal specifications, ensuring outputs satisfy semantic constraints and significantly improving reasoning and task performance.
Train models that are more performant and safer, and synthesize agents that provably adhere to formally specified constraints, including misalignment and privacy constraints.

Taken together, these advances show that formal methods provide a principled foundation for improving the utility, safety, and efficiency of frontier AI systems.

Bio

I am an Assistant Professor in the Siebel School of Computing and Data Science at the University of Illinois Urbana-Champaign (UIUC). I also co-lead the Science and Technology Working Group at the Institute of Government and Public Affairs, University of Illinois, focusing on AI governance and the future of work with AI. My research combines ideas from formal methods, machine learning, and systems research to develop systematic and theoretically principled approaches for building intelligent computing systems with formal guarantees.

Our group at UIUC has been at the forefront of advancing Trustworthy AI, pioneering state-of-the-art methods for certifying, monitoring, and synthesizing frontier AI systems with formal guarantees. Our work has been recognized through several awards, including the NSF CAREER Award, the Google Research Scholar Award, multiple Amazon Research Awards, and an Open Philanthropy research grant.