UC Berkeley professor and Fields medalist Richard Borcherds has stopped grading homework for correctness due to widespread AI use by students. He believes AI has made grading for accuracy futile as students consistently achieve high scores. Borcherds made this announcement following news of AI solving a major mathematical problem. He predicts that AI will soon surpass human capabilities in mathematics. Currently, Borcherds grades homework based solely on completion. He acknowledges this is an imperfect system but sees no immediate alternative. University attempts to use AI detection tools like Turnitin and Rumi have proven insufficient for campus-wide adoption. The responsibility for setting AI usage policies in academic work lies with individual faculty members at the university. This situation highlights the growing challenge academic institutions face in adapting to advancements in artificial intelligence. Borcherds' decision reflects a broader concern about the integrity of academic assessment in the AI era.