OpenAI has released 722 mathematical manuscripts covering 372 groups of related results, generated by an unreleased frontier AI model. The company says the work includes solutions or advances on hundreds of long-standing open problems spanning areas such as number theory and algebraic geometry.
The release follows OpenAI’s September claim that its internal model had resolved more than 100 open mathematical problems, including work related to the Navier-Stokes Millennium Prize problem. OpenAI says the new collection was produced after testing the model on roughly 4,000 problems.
The development is attracting significant attention from mathematicians while also raising questions about AI-generated research, academic credit, verification and how frontier AI companies should publish major mathematical discoveries. OpenAI has established an independent mathematics advisory group to help guide how these results are communicated and released.
