Classrooms and lecture halls have recorded a notable rise in top marks since generative AI tools became widely available.
A University of California, Berkeley analysis of more than 500,000 grades at a large Texas research university shows the share of A grades in writing- and coding-heavy courses increased by about 13 percentage points—or roughly 30% relative to the 2022 baseline—after ChatGPT’s debut.
Researchers attribute the shift primarily to students using AI for assignments rather than broad improvements in learning. The study, released this week, found larger grade increases in courses where homework carried greater weight, consistent with AI substituting for student effort.
AI Reshapes Education Landscape
The Berkeley paper, titled “Artificial Intelligence and Grade Inflation,” examined grades from 2018 to 2025 using a difference-in-differences design. Courses with more AI-exposed tasks, such as essays and coding, saw substantial rises in A grades compared with less-exposed classes. Professors awarded about 30% more A’s and fewer A-minuses and B-pluses in those subjects.
Igor Chirikov, a senior researcher at Berkeley’s Center for Studies in Higher Education and the author of the paper, stated that the results suggest students relied on generative AI to improve their assignments.
This development comes as AI adoption in education has accelerated. Global student use of AI tools jumped from 66% in 2024 to 92% in 2025, according to DemandSage data. In higher education, 86% of students now use AI as their primary research and brainstorming partner.
Teachers have also integrated the technology. A Gallup survey found 60% of U.S. K-12 public school teachers used AI tools during the 2024-2025 school year, with 32% using them at least weekly. Those who use AI weekly save an average of 5.9 hours per week—equivalent to about six extra school weeks per year.
History & Benefits of AI in Classrooms
Generative AI entered mainstream education rapidly after OpenAI released ChatGPT in late 2022. Within months, usage exploded. By the 2024-25 school year, 85% of teachers and 86% of students had used AI, per a Center for Democracy and Technology report.
Proponents point to efficiency and personalization gains. A randomized controlled trial published in Scientific Reports in 2025 found that students using an AI tutor achieved substantially higher post-test scores than those in traditional active learning, with an effect size of 0.73 to 1.3 standard deviations, in less time.
Separately, a Center for Democracy and Technology survey of middle and high school teachers found that 69% said AI tools have improved their teaching methods and skills.
AI also supports administrative relief, freeing educators for more student interaction.
Challenges and Concerns
Critics highlight risks to critical thinking and genuine skill development. A Brookings Institution report warned that AI risks currently outweigh benefits for children’s cognitive development, citing declines in content knowledge, critical thinking, and creativity among heavy users.
One student quoted in the study said, “It’s easy. You don’t need to (use) your brain,” reported NPR.
A Center for Democracy and Technology survey found that 70% of teachers worry AI weakens critical thinking and research skills, EducationWeek noted. Half of the students said using AI makes them feel less connected to teachers.
In the workforce, AI is similarly disruptive. The World Economic Forum has projected that AI will displace 92 million jobs globally by 2030 while creating 170 million new ones, for a net gain. Roles requiring AI skills command a 28% salary premium.
The technology’s rapid integration has outpaced policy development in many districts, according to RAND’s 2025 survey findings. Educators have responded with policy adjustments. Some institutions have increased in-class assessments and proctoring, or reduced the weight of homework, to verify unaided student work. Princeton, for example, reinstated proctoring requirements, per The Daily Princetonian.
As AI tools evolve, schools continue balancing potential productivity gains against concerns about learning integrity and long-term skill retention.
