Princeton University has created a new academic unit called Data and Intelligent Systems (DaIS) to speed up discovery in artificial intelligence and data science. The move, announced September 1, 2026, consolidates several existing research programs under one structure and adds new initiatives focused on AI safety and societal impact, giving faculty and students a coordinated platform for interdisciplinary work.
DaIS will be co-directed by Tom Griffiths, the Henry R. Luce Professor of Information Technology, Consciousness, and Culture of Psychology and Computer Science, and Arthur Spirling, the Class of 1987 Professor of Politics. Provost Jennifer Rexford said the unit will help Princeton "shape the course" of fast-moving AI technology while accelerating scholarship across the sciences, engineering, social sciences and the humanities.
Two pillars under one umbrella
The new unit rests on two research pillars: Princeton AI and Princeton Statistics and Data Science. These represent the evolution of the existing AI Lab and the Center for Statistics and Machine Learning. The Statistics and Data Science pillar also encompasses the Princeton Institute for Computational Science and Engineering (PICSciE), the Data-Driven Social Science Initiative, and the graduate certificate and undergraduate minor programs in statistics and machine learning.
Griffiths said the structure was designed to adapt as the fields change. "This nimble new structure will allow Princeton to continue to lead in these areas and enable us to grow in response to all of the exciting advances that are happening in the world of AI and data science today," he said.
Spirling added that placing data science on equal footing with AI research builds on Princeton's interdisciplinary strengths, positioning the University to respond "quickly but thoughtfully to the coming revolution in research."
New initiatives tackle alignment and societal impact
DaIS launches with three existing AI research initiatives - AI for Accelerating Invention, Natural and Artificial Minds, and Princeton Language and Intelligence - and will add two more this fall. The AI Alignment and Safety initiative will be led by Elad Hazan, professor of computer science and director of Google DeepMind Princeton. The Princeton Societal AI initiative will be co-led by Janet Vertesi, associate professor of sociology, and Matt Jones, the Smith Family Professor of History.
"We are not just thinking about how we innovate on the core technologies, but how we do so in a thoughtful and responsible way that grapples with the very real impact these technologies are having on our society," Griffiths said. "We want to make sure these developing systems are beneficial for humans in the long run."
Vertesi described the practical advantage of housing these conversations under one roof. "We have an interdisciplinary research team of humanists, social scientists and computer scientists across multiple departments," she said. "What is exciting about DaIS is that we will have a place for these interdisciplinary conversations under a single umbrella."
Bridging computational fields
Michael E. Mueller, the Donald R. Dixon '69 and Elizabeth W. Dixon Professor of Mechanical and Aerospace Engineering and former director of PICSciE, pointed to the blurring lines between computational science, data science and AI. "Clear boundaries between computational science and engineering, data science and AI are no longer apparent," he said. "Anything that we can do to better coordinate this work and facilitate an exchange of ideas will only accelerate progress."
The launch of DaIS is the latest step in Princeton's broader AI strategy, which includes the University's partnership with the State of New Jersey, Microsoft and CoreWeave in the NJ AI Hub. For researchers navigating the intersection of computation and domain science, the consolidation signals a shift toward unified support structures. The AI Learning Path for Research Scientists addresses similar needs for professionals building interdisciplinary AI skills.
Why this matters for science and research professionals
DaIS is not a funding announcement - it is a structural signal. Princeton is betting that the fastest progress in AI will come from collapsing the administrative walls between statistics, machine learning, computational science, and the humanities. For researchers in academia and industry, the takeaway is practical: institutions are reorganizing around the reality that AI work now requires fluency across methods and domains. Units like DaIS create the kind of environment where a political scientist and a computer scientist can co-lead a major research platform, and where safety research sits alongside core technology development from day one. The model may influence how other universities and research labs structure their own AI for Science & Research efforts in the coming years.
Your membership also unlocks:
