Cambridge and King's Researchers Document Gaps Between OpenAI Proof Text and Lean Code

Researchers at the University of Cambridge and King's College London documented at least two discrepancies between OpenAI's natural-language proof and Lean code for a Navier-Stokes-derived problem. Reported October 8, 2026, amid OpenAI's release of hundreds of claimed solutions, of which only 10 of 719 manuscripts included chain of thought.