Mathematicians Criticize OpenAI’s 300+ Mathematical Proofs as Low Quality

15 Oct 2026, 1:05 pm · 2d ago · 1 min read · OfficeChai

Prominent mathematicians, including Sophie Huiberts and Fields Medalist Terence Tao, have criticized OpenAI's recently published collection of more than 300 mathematical proofs. Shared publicly on October 6, the automated proofs were described by Huiberts as low quality and comparable to easy rejects in standard academic peer review. The critique doubles down on scrutiny surrounding frontier AI reasoning capabilities, challenging claims that modern large language models can perform autonomous, rigorous mathematical research. The pushback highlights persistent gaps between automated pattern generation and verifiable, high-level mathematical problem solving required by academic and enterprise researchers globally.