NSF Boosts MIT AI-Physics Institute Funding to $5M Annually

The National Science Foundation has renewed support for the MIT-led Institute for Artificial Intelligence and Fundamental Interactions (IAIFI) with increased annual funding of $4.98 million, up from $4 million, for an additional five years through 2031.

The renewal validates IAIFI's central thesis: that artificial intelligence can open new pathways in physics research, while physics principles can produce more reliable and interpretable AI systems. Since its 2020 launch as part of the NSF's National Artificial Intelligence Research Institutes program, IAIFI has built an interdisciplinary research community spanning MIT, Harvard, Northeastern, Tufts, and Boston universities.

Research bridging two domains

IAIFI's work demonstrates bidirectional benefits between the fields. In particle physics, researchers have developed AI techniques to process the enormous data volumes from the Large Hadron Collider in real-time. In nuclear physics, AI-based generative methods are modeling quark and gluon interactions in lattice quantum chromodynamics. Astrophysics applications include discovering new cosmic phenomena and improving sensitivity at the MIT-led LIGO gravitational-wave experiment.

The reverse direction proves equally productive. IAIFI researchers are embedding physics knowledge—symmetries, geometric structures, exactness guarantees, and statistical methodologies—directly into neural network architectures. The result: AI systems that are more data-efficient, reliable, and interpretable than conventional approaches.

"AI has begun to transform how physicists tackle some of the field's most challenging problems," says Mike Williams, IAIFI's interim director and an MIT physics professor. "More importantly, it is starting to expand the frontier of what problems we can realistically address."

Training 'centaur scientists'

The institute's fellowship program supports early-career researchers working at the physics-AI intersection, pairing each fellow with mentors in both domains. Of eight completed fellows, three have secured faculty positions while others joined leading AI companies or startups.

IAIFI's annual PhD Summer School has become a competitive draw, receiving nearly 600 applications for approximately 100 in-person spots in 2026, with 300 additional virtual participants expected. At MIT specifically, the institute helped shape a new interdisciplinary PhD program in physics, statistics, and data science that has awarded 20 doctoral degrees since 2021.

Why it matters

The funding increase and five-year extension signal federal confidence in cross-disciplinary AI research models at a time when both physics and AI face fundamental questions about interpretability, reliability, and the limits of current methods. IAIFI's approach—embedding domain expertise into AI systems rather than treating machine learning as a black box—offers a template for other scientific fields seeking to leverage AI while maintaining rigor. The institute's success in training researchers fluent in both disciplines addresses a critical talent gap as AI becomes central to scientific discovery.

"The first phase of IAIFI established the model: interdisciplinary research, early-career talent, and a dynamic community, organized around the idea that AI and physics make each other stronger," says Jesse Thaler, IAIFI's director and an MIT physics professor currently on sabbatical. "Now we have the foundation—and the entrepreneurial spirit of our centaur scientists—to push that model into new territory."

Details of the renewed funding were first reported by MIT News.