A new study from the University of California, Berkeley, released Wednesday, shows that the share of A’s in college classes heavy on writing and coding has grown more significantly than in other classes since ChatGPT’s debut. Professors teaching AI-exposed classes gave out about 30% more A’s and fewer A-minus and B-plus grades. The results suggest that students have relied on generative AI to do better in their studies, not that these classes of students are learning more, says Igor Chirikov, a senior researcher at Berkeley’s Center for Studies in Higher Education and the paper’s author. Chirikov pulled data on more than half a million grades from 2018 to 2025 at a large public university in Texas. There was little difference between AI-exposed and less-exposed classes through 2022; afterward, “A” grades shot up higher in the more AI-exposed courses. In 2023, 37% of employers used grade-point average in hiring, according to the National Association of Colleges and Employers, a figure that has ticked back up to 42%. Employers including Barclays and Morgan Stanley have GPA minimums for some internship roles. The concern is that AI is making that evaluation method less reliable.