Mathematical Foundations of Artificial Intelligence
This grant aims to address foundational gaps in AI through interdisciplinary research on mathematical and theoretical foundations for sustainable, trustworthy, and socially responsible AI.
Browse similar grants
Funder: U.S. National Science Foundation
Due Dates: October 10, 2024 (Full proposal) | October 10, 2025 (Full proposal) | October 9, 2026 (Full proposal)
Funding Amounts: $500,000–$1.5 million total per award over 36 months; up to $8.5 million for new awards annually
Summary: Supports interdisciplinary research on the mathematical and theoretical foundations of artificial intelligence.
Key Information: PIs, co-PIs, and senior/key personnel must hold a qualifying appointment at a U.S.-based campus of an eligible organization.
Description
The U.S. National Science Foundation’s Mathematical Foundations of Artificial Intelligence program supports collaborations focused on difficult theoretical questions about AI’s capabilities, limitations, and emerging properties. It is jointly sponsored by NSF directorates covering mathematical and physical sciences, computer and information science and engineering, engineering, and social, behavioral and economic sciences.
Projects should advance mathematical understanding of current or emerging AI systems or develop mathematically grounded principles for designing and analyzing them. Goals include rigorous validation of machine-learning algorithms and predictions, provably reliable AI, and explainable and interpretable models. Teams should bring together complementary expertise appropriate to the project and address how their mathematical advances relate to AI applications or observed behavior.
