Terence Tao, along with 24 other Fields Medalists, collectively issued a statement titled "Significant Inaccuracies of AI in Mathematical Research." In this statement, they acknowledge the significant potential of AI in advancing mathematics. However, they raise concerns regarding the rapid, large-scale generation of solutions by AI, which often lack standardized structuring. This makes it challenging to discern between genuinely innovative methods, novel concepts, and mere references to prior research. They fear that this could erode the fertile environment for fostering new ideas and disrupt the continuity of human intellectual legacy among mathematicians. Given the pressing nature of the issue, the statement was released without undergoing a full consultation process.

The day before the statement's release, a preprint surfaced, marked as having been discovered and proven by the Odin Automatic AI Research Agent. It claimed to have resolved the Komlós conjecture, a problem that had lingered in discrepancy theory for four decades. The solution proposed a constant of 3√(2π) and suggested implications for the Beck–Fiala conjecture. The paper's authors were three biostatistics scholars who had previously used Odin to assert a solution to the Talagrand convolution conjecture, earning recognition from their peers.

However, Odin lacks public records, and the paper only indicates that AI was responsible for discovering the proof, with the authors arranging it into a written format. Currently, there is a flood of mathematical results claiming AI involvement, which has far outpaced the peer-review process, undermining the foundations of traditional review mechanisms. How the mathematical community can harmoniously coexist with AI's computational capabilities has emerged as a new challenge to be addressed.