More students fail computer science classes as AI reliance grows
What's the story
The University of California, Berkeley, has seen a spike in failing grades in its computer science classes. The trend was particularly prominent during the Spring 2026 semester. The percentage of students who failed CS 10 and CS 61A courses was significantly higher than previous semesters, according to data from Berkeleytime. Professors have attributed the surge in failing grades to a number of factors, including students' increasing dependence on artificial intelligence (AI) tools.
Cause analysis
More students receive F's
A lack of mathematical preparedness and understaffing were also cited as potential reasons for this trend. In Spring 2026, 35.3% of CS 10 students and 10.6% of CS 61A students received F's, far exceeding the department's grading guidelines, which expect only about 7% to receive D's or F's in lower division courses like these two.
Academic integrity
Garcia blames academic dishonesty, lack of math skills
Dan Garcia, a teaching professor at UC Berkeley, said that the high failing rates are largely due to an increase in academic dishonesty. He said this is mainly because students are using large language models (LLMs) for their work. Garcia also noted that many students lack basic skills needed for these courses.
Grading transparency
Garcia opposes limiting A's to a fraction of students
Garcia suggested that instructors should be transparent about their grading criteria and give students multiple opportunities to meet them. He said he is against the practice of limiting A's to a fraction of students, as done by Harvard. "I think you should have clear standards for what an A means, and then give tons of opportunity for people ... to get to that A bar without lowering the standard," Garcia said.
Testing debate
Professors call for reinstatement of standardized testing scores
Both Garcia and Gireeja Ranade, an associate teaching professor at UC Berkeley, have signed a petition with over 1,300 other faculty members. The petition calls for the reinstatement of ACT and SAT standardized testing scores for STEM admissions in the UC system. The professors argue that students' mathematical preparation has declined significantly over the years.
Class dynamics
Other factors contributing to high F rates
Ranade had to remove a final project from her EECS 127 class due to understaffing. The course saw a 16.8% F rate, much higher than the typical 5% of D's and F's for an upper division course. Both professors have also noticed a decline in student engagement during classes with low office hour attendance despite repeated encouragements from Ranade and her TAs.