Simons Foundation Partners With Anthropic and OpenAI on $15 Million AI Initiative for Mathematics
The Simons Foundation has announced a $15 million initiative in partnership with Anthropic and OpenAI to help mathematicians evaluate, adopt and establish standards for artificial intelligence tools as the technology becomes increasingly integrated into mathematical research.
The initiative will combine funding and resources from the three organizations to support workshops, research collaborations and community-led programs focused on the opportunities and challenges associated with AI-assisted mathematics.
The Simons Foundation will administer and award the funding, including through an open call for proposals designed to engage researchers across the mathematics community.
The program will support efforts to assess recent AI-assisted mathematical results, provide training on emerging tools, encourage collaboration and develop shared approaches to the responsible use of AI in research.
David Spergel, President of the Simons Foundation, said the initiative is intended to ensure that mathematicians play a central role in determining how artificial intelligence is incorporated into their field.
“As artificial intelligence rapidly reshapes the research landscape, we want to help the mathematical community come together and forge the best path forward,” Spergel said.
“Now is the time to look at mathematics as a whole — not as a collection of proofs and theorems, but as a collective understanding driven by human curiosity.”
Under the partnership, Anthropic and OpenAI will also provide AI tools and computing resources to researchers supported by the Simons Foundation.
Those resources will extend beyond mathematics to include researchers working in other areas of basic science.
The initiative comes as advances in artificial intelligence are creating new possibilities for mathematical exploration, including automated reasoning, proof development and the verification of complex mathematical arguments.
AI-assisted mathematical research also raises questions about the reliability of generated results, the standards required for verification and the appropriate role of human researchers in evaluating machine-generated work.
The Simons Foundation’s program is designed to provide opportunities for mathematicians to examine these issues collectively rather than relying exclusively on standards developed by technology companies.
Workshops and other funded activities will allow researchers to share practical experiences, assess the capabilities and limitations of emerging systems and discuss how AI tools should be integrated into established mathematical research practices.
The foundation emphasized that the initiative will be community-driven, with mathematicians helping determine which questions, tools and practices warrant further attention.
“This is a moment for the mathematics community to come together,” Spergel said.
“The community should be driving discussions of how these tools are evaluated, adopted and used, not any single company or institution. The goal of the program is to support the mathematics community.”
The initiative builds on the Simons Foundation’s longstanding support for mathematics and computational research.
The foundation was an early supporter of machine learning applications in science and has funded the development of Lean, a proof assistant used to formally express and verify mathematical arguments.
Lean enables mathematical statements and proofs to be represented in a form that can be checked by computer, providing a framework for verifying the logical correctness of mathematical reasoning.
Such systems have become increasingly relevant as researchers explore the use of AI to generate mathematical proofs and assist with complex reasoning tasks.
Formal verification tools can help researchers distinguish between plausible AI-generated arguments and proofs that satisfy rigorous logical requirements.
The foundation has also organized workshops, lectures and research collaborations examining the relationship between artificial intelligence and mathematics.
The new partnership expands those efforts by combining financial support with access to AI technology and computing infrastructure.
Although the organizations announced a combined commitment of $15 million in funding and resources, they did not specify how much each partner will contribute or how the total will be divided between grants, computing resources and other support.
Details regarding application deadlines, individual award amounts and the selection process were not included in the announcement.
Founded in 1994 by Jim and Marilyn Simons, the New York City-based Simons Foundation supports research in mathematics and the basic sciences through grants, internal research programs and public engagement.
Its funding areas include mathematics and physical sciences, life sciences, neuroscience, autism science and initiatives connecting science with society and culture.
The foundation also operates the Flatiron Institute, an in-house research organization focused on developing and applying computational methods to advance scientific discovery.
Through the new initiative, the Simons Foundation, Anthropic and OpenAI aim to provide researchers with resources to explore AI-assisted mathematics while supporting community-led evaluation and standards development.
The partnership represents a coordinated effort to help mathematicians navigate technological changes while maintaining the field’s emphasis on rigorous verification, independent inquiry and human understanding.
