Simons Foundation, OpenAI and Anthropic put $15M behind mathematicians wrestling with AI proofs

What's happening? On 15 October 2026 the Simons Foundation announced a partnership with OpenAI and Anthropic committing a combined $15 million in funding and resources to help mathematicians evaluate AI-assisted results, learn new tools and shape collective standards for AI use in the field. The foundation will manage and award the money, including through an open call for proposals for workshops and related programmes.

Verified: OpenAI says it is contributing $5 million to the initiative and will also provide model access and compute to Simons-supported researchers. Anthropic is likewise contributing funding and tools under the joint package. Support is described as extending beyond pure mathematics into basic science. The foundation, an early backer of machine-learning-for-science work including the Lean proof assistant, frames the programme as community-driven rather than a lab-run peer-review substitute.

Context: The announcement lands days after OpenAI publicised hundreds of experimental AI-generated maths results — coverage that drew sharp pushback from mathematicians who said a mass dump of machine proofs overwhelms traditional verification and collaboration norms. The $15M purse is less a celebration of solved conjectures than an admission that the research community needs infrastructure to check, interpret and build on what models now emit at volume.

Why it matters: Frontier labs are starting to treat the scientific fields they disrupt as stakeholders who need funded capacity, not only as benchmark scoreboards. Watch how awards are allocated, whether independent verification norms emerge, and whether other labs join — the initiative does not by itself resolve disputes over priority, credit or the quality of AI-generated proofs.

Sources: Simons Foundation (15 Oct 2026); OpenAI mathematics update; The Verge; The Independent.