The Simons Foundation has partnered with OpenAI and Anthropic to launch a $15 million initiative aimed at helping mathematicians evaluate artificial intelligence-generated research, develop shared standards for AI use and explore its applications in mathematical discovery.
Announced on October 15, the initiative will combine financial support with AI tools and computing resources. The Simons Foundation will manage and distribute the funding through an open call for proposals and other programmes designed to support the mathematics research community.
The initiative will fund workshops, conferences and collaborative activities where researchers can examine recent AI-assisted mathematical findings, assess their reliability and explore methods for incorporating AI into research practices.
OpenAI has separately confirmed a $5 million contribution to the programme. The company said the funding would support mathematicians in examining, understanding and building on results produced by AI, including mathematical findings recently shared by OpenAI.
Both OpenAI and Anthropic will provide access to AI models and computing resources for researchers supported by the Simons Foundation. The assistance will also extend to other areas of basic scientific research.
The programme comes as AI developers increasingly demonstrate advances in mathematical reasoning, problem-solving and proof generation. These developments have prompted discussions within the research community about how AI-generated results should be verified and incorporated into established scientific processes.
Simons Foundation President David Spergel said the initiative would bring mathematicians together to determine how emerging AI capabilities should be evaluated and adopted.
“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.
The foundation has emphasised that decisions concerning the use of AI in mathematics should be guided by researchers rather than individual technology companies or institutions.
The initiative builds on the Simons Foundation's existing investments in mathematics and machine learning for scientific research. Its previous funding activities include support for the development of Lean, a proof assistant used to formalise mathematical statements and verify proofs.
Such tools allow researchers to examine mathematical arguments through formal verification processes, providing an additional mechanism for assessing results developed with AI assistance.
The foundation has also organised lectures, workshops and research collaborations examining the relationship between artificial intelligence and mathematics.
According to the organisation, the new programme will encourage researchers to exchange findings, develop common evaluation practices and establish standards for using AI tools in mathematical work.
While AI systems are increasingly capable of generating mathematical arguments, questions remain around verifying their correctness, interpreting their methods and determining their contribution to scientific understanding.
The foundation has not announced a detailed timeline for distributing the funding or the number of research projects expected to receive support.
The partnership brings together two major AI developers and an independent research foundation, with the stated objective of supporting researcher-led evaluation and adoption of AI in mathematics.
Disclaimer: This article may include information derived from interviews, press releases, public statements, research, company communications and other publicly available or third-party sources. Such material may be summarised, paraphrased or contextualised for journalistic and editorial purposes. All rights in third-party content remain with their respective owners.