University of California, San Diego

The University of California, San Diego (UCSD) is a public research university founded in 1960 and located in La Jolla, San Diego, California. It is part of the University of California system and consistently ranks among the world's leading research universities, with its Computer Science and Engineering department ranked third nationally for CS programs and sixth for AI (CSRankings.org). The university enrolls approximately 46,000 students and employs over 41,000 people, making it the largest employer headquartered in San Diego County. Total revenues for FY 2023-24 were $9.1 billion, with $1.73 billion in sponsored research funding for FY 2024-25. AI safety-relevant research at UCSD is conducted across several labs and institutes. Professor Benjamin Bergen in the Department of Cognitive Science leads the Language and Cognition Lab, which received a $470,731 grant from Open Philanthropy in July 2024 to evaluate AI persuasiveness - specifically studying LLMs' ability to change minds, foster cooperation or defection in game scenarios, and negotiate for hidden goals through freeform conversation. The lab also conducts research on LLM deception and manipulation risks. Assistant Professor Lily Weng at the Halicioglu Data Science Institute (HDSI) leads the Trustworthy ML Lab, which received NSF CISE/IIS Core Program funding (Award No. 2430539) for research on interpretable neural networks with robustness guarantees. The work aims to make AI decision-making transparent and reliable for critical domains including healthcare, autonomous driving, and criminal justice. Professor Sylvia Herbert leads the Safe Autonomous Systems Lab, which develops methodologies for analyzing the safety of autonomous systems, including reach-avoid safety filters and Hamilton-Jacobi reachability analysis. Professor Farinaz Koushanfar co-directs the Center for Machine Intelligence, Computing, and Security (MICS), which conducts research on AI security. At the institute level, UCSD participates in TILOS (Institute for Learning-enabled Optimization at Scale), part of a $220 million NSF National AI Research Institutes initiative, and EnCORE, an NSF-funded institute addressing responsible computing and data science foundations.