University of California, San Diego
UC San Diego (UCSD) is a top-ranked public research university based in La Jolla, California, with extensive AI research programs across multiple departments and institutes. AI safety-relevant work includes Benjamin Bergen's Open Philanthropy-funded research on evaluating the persuasive capabilities of large language models, Lily Weng's NSF-funded Trustworthy ML Lab focused on interpretable and robust deep neural networks, and Sylvia Herbert's Safe Autonomous Systems Lab researching safety guarantees for autonomous systems. UCSD's Halicioglu Data Science Institute and the TILOS and EnCORE NSF AI Research Institutes also contribute to responsible computing and trustworthy AI research.
Funding Details
- Annual Budget
- $9,100,000,000
- Monthly Burn Rate
- -
- Current Runway
- -
- Funding Goal
- -
- Funding Raised to Date
- -
- Fiscal Sponsor
- -
Theory of Change
As a research university, UCSD's contribution to AI safety operates primarily through knowledge generation and talent development. Individual researchers and labs produce technical work on evaluating dangerous AI capabilities (persuasion, deception), building interpretable and robust AI systems, and ensuring safety in autonomous systems. By advancing understanding of AI risks and building tools to make AI systems more transparent and controllable, this research feeds into the broader AI safety ecosystem. Funded work like Bergen's persuasion evaluation directly informs policymakers and developers about emergent dangerous capabilities in frontier models. Trustworthy ML research provides developers with frameworks for building safer systems. Graduate student training creates a pipeline of researchers capable of working on AI safety problems across academia and industry.
Grants Received
from Open Philanthropy
Projects
No linked projects.
People
No linked people.
Discussion
Sign in to join the discussion.
No comments yet. Be the first to share your thoughts.
Details
- Last Updated
- Apr 2, 2026, 9:54 PM UTC
- Created
- Mar 20, 2026, 2:34 AM UTC