The Pulse
Simons Foundation Leads $15M AI-and-Math Partnership
The Simons Foundation, Anthropic and OpenAI are committing $15 million in combined funding and resources to help mathematicians assess AI-assisted work and develop shared standards. The foundation will manage awards, including through an op
AI.info Team · Written with AI ·
“The community should be driving discussions of how these tools are evaluated, adopted and used, not any single company or institution.”
David Spergel, president, Simons Foundation
The Simons Foundation, Anthropic and OpenAI announced a $15 million partnership on October 15 to help mathematicians respond to the growing use of artificial intelligence in research. The foundation will manage and award the support, including through an open call for proposals, while Anthropic and OpenAI will provide AI tools and computing resources to researchers it supports.
The program is intended to fund workshops and other initiatives where researchers can evaluate AI-assisted results, learn to use new tools, share perspectives and develop collective standards for AI use. The announcement does not specify how the $15 million is divided between the partners or between funding and other resources.
Simons will decide who receives support
Giving the foundation control over the awards places grant selection outside the two AI companies, even as they contribute resources to the effort. The foundation says the open call will support community-driven work, but its announcement does not give an application deadline, eligibility rules or a schedule for awards.
Spergel framed the initiative as a chance for mathematicians to shape how AI enters their discipline, rather than leaving decisions to technology companies or a single institution. The stated goals range from checking recent AI-assisted results to building shared expectations for how researchers use the tools.
The $15 million includes funding and resources
The headline figure covers “combined funding and resources” from the three partners. The announcement does not break out each partner’s contribution, identify how much will be distributed as grants, or put a separate dollar value on the AI tools and computing resources Anthropic and OpenAI will provide.
Those tools and resources are not limited to mathematics: the foundation says they will also be available to researchers it supports in basic science. That gives the partnership a broader reach than its central focus on mathematicians, though the announcement does not describe which basic-science projects will qualify or how access will work.
Lean is part of the foundation’s AI-math history
The Simons Foundation points to earlier support for machine learning in science, including funding the development of Lean, a proof assistant used to formalize mathematical arguments. Formal proof systems can let researchers check whether a proof follows from stated assumptions, a task that has drawn renewed attention as AI systems produce mathematical claims.
The foundation also says it has convened workshops, lectures and collaborations on AI’s role in mathematics. The new initiative extends that work with funding and access to tools, while leaving the research community to propose the specific activities and standards it wants to pursue.
Details mathematicians still need
The announcement establishes the partnership’s total commitment and broad purpose, but leaves practical questions unanswered. It does not say when the call will open, who may apply, how proposals will be reviewed or when researchers can expect decisions.
Nor does it explain how the partners will distinguish support for evaluating AI-generated mathematics from support for developing or using the systems themselves. For now, the concrete commitments are a foundation-run award process, an open call to come, and AI tools and computing resources from Anthropic and OpenAI for foundation-supported researchers in mathematics and basic science.