OpenAI commits $5M to $15M Simons Foundation-led math and AI research programme on October 15

OpenAI on October 15 announced $5M to a $15M Simons Foundation-led programme, with Anthropic also funding, supporting mathematicians evaluating the quality, interpretability and reproducibility of AI-generated mathematical research. The programme, led by the Simons Foundation, will establish a dedicated math-and-AI research group, organise systematic reviews by mathematicians of frontier-model-generated proofs, theorems and conjectures, publish open evaluation reports and benchmark datasets, and partner with university math departments to run AI-mathematics courses. Anthropic simultaneously announced $5M to the same programme. Two leading frontier labs jointly funding a single mathematics-and-AI programme is unusual in public reporting, reflecting the rapid expansion at the intersection of AI and pure mathematics and shared concern over the interpretability of model mathematical capability.