OpenAI's math model claims 100 open problems

OpenAI says its internal model has resolved more than 100 open mathematics problems, as it forms a nine-member independent advisory group to review how those results are presented and released. The Institute for Advanced Study’s group charter says members are unpaid and independent but have no decision-making power at an AI company; TechCrunch’s report notes the claim has drawn scrutiny after OpenAI’s Navier–Stokes announcement. For scientific-AI builders, file this as a provenance and verification signal—not a published proof—and pair any model-assisted result with the archive’s earlier OpenAI mathematics analysis.