On September 11, 2026, MIT professor Dor Minzer began receiving messages about a rumored OpenAI proof of the unique games conjecture, a famous open problem in computational complexity. Minzer had not solved the conjecture, but he and his graduate students Yumou Fei and Shuo Wang had proved a related milestone result and were still writing it up. Fearing their work would be overshadowed, they rushed to finish and posted a 95-page paper three days later with a note saying the manuscript was mathematically complete but not in the shape they wished to share.
The unique games conjecture concerns constraint satisfaction problems, such as coloring the nodes of a graph so that each edge's rule is obeyed. The conjecture states that even finding a poor approximation can be extremely hard in some cases. Researchers have used the conjecture to study other problems, and a proof would imply that current methods for many unrelated problems cannot be improved.
On October 6, OpenAI announced a proof of the unique games conjecture, along with 376 other results across mathematics. The announcement drew a mix of alarm and curiosity from researchers. Minzer's result, by contrast, was met with praise: Carnegie Mellon's Ryan O'Donnell called it 'another truly great one.'