Mathematicians say OpenAI doesn't understand its own Navier-Stokes proof

OpenAI released a 166-page document attempting to solve the Navier-Stokes existence and smoothness problem, but academic experts report that the underlying methodology is inconsistent with the model's own description. Because large language models function as opaque systems, researchers are struggling to verify the validity of the logic or determine if the AI truly reached a mathematical breakthrough. This incident highlights the ongoing difficulty of auditing AI-generated scientific claims when the internal reasoning processes remain largely inaccessible.