OpenAI announced on October 6, 2026 that it is releasing a broad range of new mathematical results produced by an internal frontier model. The results are being shared via a GitHub repository that includes protocols for paper revisions and citations, following consultations with the independent Advisory Group on Mathematics and Artificial Intelligence at the Institute for Advanced Study to develop best practices for sharing AI-generated mathematical work with the research community.
This image is a promotional graphic from OpenAI's announcement page, visually representing the theme of sharing AI-generated mathematical progress with the research community.
The repository contains formalizations of many proofs in Lean, a programming language that allows computer verification of mathematical proofs. In addition, OpenAI is providing ten summaries of the model’s reasoning process for these results, estimates of the computational resources expended measured in ChatGPT Pro usage equivalents, and statistics on the number of problems attempted during discovery. According to the release, the average result required computational resources equivalent to roughly three hours of ChatGPT Pro thinking time.
This image depicts an art card emphasizing mathematical breakthroughs, aligning with the article's discussion of new AI-generated mathematical results.
OpenAI states that this release aims to promote scientific transparency and openness while pushing the frontier of human knowledge in mathematics. The organization plans to fund workshops, conferences, and special programs focused on understanding major results produced by AI systems. It also indicates it is working toward responsibly releasing the model that generated these results, emphasizing the importance of continuing to evaluate internal frontier models on mathematics and other sciences to develop tools that can advance these fields, and committing to update its standards for future disclosures based on community feedback.