Princeton engineers uncovered a universal weakness in AI chatbots that makes it possible to bypass their safety guardrails and unlock malicious uses with just a few lines of code. Illustration by Alaina O’Regan
Nimble, high-intensity research teams at Princeton collaborate across disciplines to accelerate discovery in artificial intelligence without the logistical barriers that have traditionally slowed universities down. Engineers, scientists, humanists and policy experts who are pre-eminent in their fields come together in a deeply interdisciplinary approach to problem solving.
Featured Faculty
Developing conceptual understanding and better design methods for today's AI models.
Understanding what governments and their citizens can expect from, and how they can adapt to, AI in politics.
Building AI methods to decode genomes and diagnose and treat complex diseases.