Mathematical Foundations of Artificial Intelligence (MFAI)

The NSF Mathematical Foundations of Artificial Intelligence (MFAI) program supports research collaborations focused on the most challenging mathematical and theoretical questions aimed at understanding the capabilities, limitations, and emerging properties of AI methods. The program aims to establish a fundamental mathematical understanding of factors determining capabilities and limitations of current and emerging AI systems, including foundation models, generative models, deep learning, statistical learning, federated learning, and other evolving paradigms. Research should focus on developing mathematically grounded design and analysis principles for current and next generations of AI, rigorous approaches for characterizing and validating machine learning algorithms and their predictions, and research enabling provably reliable, translational, general-purpose AI systems and algorithms. The program is jointly sponsored by NSF Directorates for Mathematical and Physical Sciences (MPS), Computer and Information Science and Engineering (CISE), Engineering (ENG), and Social, Behavioral and Economic Sciences (SBE). PI teams should collectively possess appropriate expertise drawn from mathematics/statistics, computer science, engineering, and social and behavioral sciences. The overall goal is to establish innovative and principled design and analysis approaches for AI technology using creative yet theoretically grounded mathematical and statistical frameworks, yielding explainable and interpretable models that can enable sustainable, socially responsible, and trustworthy AI.