AI Grade Inflation Surges 30% at Universities as Students Learning Less, Study Reveals

AI Transforms Academic Performance Across Universities

The widespread use of generative AI is fundamentally reshaping how students perform academically, according to a working paper from the University of California, Berkeley

University of California, Berkeley

. UC Berkeley senior researcher Igor Chirikov analyzed over 500,000 student-course enrollments across 84 departments at a large Texas university from 2018 to 2025, uncovering a troubling pattern: students are achieving significantly higher grades while potentially acquiring fewer actual skills. The research identifies three distinct ways students deploy these tools—augmentation for research support, reinstatement of AI-based tasks, or complete displacement where AI automates work like essay writing. Only the first two methods correlate with genuine learning, yet displacement remains widespread in unsupervised environments.

AI-Exposed Courses See Dramatic Grade Increases

Since ChatGPT

ChatGPT

launched in 2022, AI-exposed courses experienced a 30 percent increase in "A" grades, with the most pronounced effects appearing in classes involving writing and coding tasks

writing and coding tasks

. "We have a C student who is now an A student," Chirikov told Axios

Chirikov told Axios

, emphasizing this isn't merely marginal improvement. Meanwhile, courses less susceptible to AI assistance—such as sculpture and lab-based work—saw grades remain flat, providing clear evidence that AI-assisted cheating drives these academic gains. The pattern becomes especially pronounced in courses where take-home assignments carry significant weight compared to proctored exams, oral presentations, or in-class discussions.

Universities Struggle to Address AI-Fueled Academic Changes

Four years into the AI era, universities have yet to develop effective responses to grade inflation. Chirikov deliberately avoided naming the Texas research university in his UC Berkeley study

UC Berkeley study

, explaining "I don't want to single out one university, just because I believe it's not specific to that particular university. It's something that's happening across the higher ed sector." The challenge extends beyond student behavior—faculty sometimes face incentives to grade more leniently since student evaluations often influence promotions. At Princeton

Princeton

, where roughly 30% of seniors admitted to cheating mostly via generative AI in a recent survey, faculty voted this week to overturn a 133-year-old honor code

honor code

that allowed students to take in-person exams without faculty proctoring.

Workforce Competency Concerns Mount for AI-Friendly Subjects

The implications extend far beyond campus boundaries. "If AI displaces skill-building tasks during learning, students may graduate with weaker capabilities in precisely the domains where AI is strongest, reinforcing a feedback loop between AI in education and AI in production that could accelerate automation," Chirikov writes

Chirikov writes

. This creates a paradox: graduates enter workplaces unable to perform core duties, driving increased workplace AI dependence and potentially accelerating job automation. Employers will struggle to identify strong graduates

strong graduates

when inflated GPAs no longer signal actual competency, making hiring decisions increasingly difficult in a job market already strained by AI displacement.

Future Approaches to AI in Academic Settings

Chirikov acknowledges there's no "silver bullet" to address AI and grade inflation at universities. "We need to be creative and think of AI-integrated assignments

AI-integrated assignments

, and that students can use [LLMs], but they should properly document that," he told Axios

he told Axios

. "That's not an easy process, but we definitely should invest in that more than we do right now." Some professors have already implemented creative countermeasures, requiring handwritten or oral exams to verify authentic student knowledge. The research suggests that while AI exacerbates existing grade inflation trends dating to the early 2000s, the speed and scale of change demands immediate attention. What remains uncertain is whether academic institutions can adapt quickly enough to preserve educational integrity while preparing students for an AI-integrated professional world.