Princeton University

Princeton University, founded in 1746 and located in Princeton, New Jersey, is one of the oldest and most prestigious research universities in the United States. With an endowment of approximately $34.1 billion and an annual operating budget of roughly $3.3 billion, it supports a wide range of AI safety-relevant research across multiple centers and initiatives. The Center for Information Technology Policy (CITP) is Princeton's primary hub for technology policy research, housed within the School of Public and International Affairs. Under the direction of Prof. Arvind Narayanan (appointed director in 2023), CITP has become a significant contributor to global AI safety discourse. CITP experts including Narayanan, Alondra Nelson, and Sayash Kapoor contributed to both the first International AI Safety Report (January 2025) and its 2026 update. CITP research covers AI governance, biosafety risks in AI foundation models, and the evaluation of open foundation models. The center runs the Princeton AI Dialogues, which connects researchers with Congress and federal agencies including Commerce, Defense, and Homeland Security through briefings and roundtables. The Princeton Language and Intelligence (PLI) initiative, launched in September 2023 and led by Prof. Sanjeev Arora, focuses on developing fundamental understanding of large AI models, enabling their application across disciplines, and studying their societal and ethical implications. In its inaugural round of seed grants, PLI awarded $798,000 to 14 interdisciplinary research projects. PLI has access to a $10 million computing budget and infrastructure including 300 Nvidia H100 GPUs. Princeton AI Alignment (PAIA) is a campus-based community supporting Princeton students and researchers working on AI alignment. It runs an 8-week introductory seminar, an advanced fellowship covering state-of-the-art alignment research, mentored research opportunities, and hosts workshops. PAIA alumni have joined leading AI safety organizations including OpenAI and Anthropic. The Princeton Dialogues on AI and Ethics, a collaboration between CITP and the University Center for Human Values active since 2017-18, develops interdisciplinary reasoning tools for practitioners and policymakers on AI ethics and fairness. Princeton has received multiple Open Philanthropy grants for AI safety-relevant research: $100,000 for scalable oversight research (Prof. Jason Lee), $863,143 for an AI R&D capabilities benchmark (Prof. Arvind Narayanan), and over $1 million for a software engineering LLM benchmark (Prof. Karthik Narasimhan).