Concerns Over High Failure Rates in UC Berkeley Computer Science Courses

With alarming failure rates in UC Berkeley’s computer science courses, professors are urging the school’s administration to bring back standardized testing.

An analysis by the student newspaper revealed that this spring’s failure rate surpassed that of the previous semester.

More than 35% of students failed the introductory computer science course, typically described as a “gentle but thorough introduction to computer science.” This contrasts sharply with the usual failure rate of about 7%. Other courses also showed a notable rise in failure scores.

Professor Dan Garcia, who teaches two of these courses, expressed that many students lean too heavily on artificial intelligence tools to navigate their studies. Some AI models, such as Claude from Anthropic, are well-known for their coding capabilities.

Garcia noted that nearly 30 students were caught cheating on the take-home exam in the CS 10 class alone.

“In some instances, students depend on the LLM too much, and then they’re unprepared for exams,” he explained.

He also highlighted a trend of students being mathematically underprepared. In fact, he was one of 1,300 UC faculty members who recently stated in a letter that they were effectively teaching “middle school” math in calculus and other subjects.

Faculty disapproval has been directed at a 2020 decision by the University of California Board of Regents that eliminated the requirement for SAT and ACT scores for admissions, following legal arguments from advocates for low-income students that deemed these tests “racist.”

Another professor, Girija Ranade, recounted a student mentioning that the linear algebra class operates under an “open internet, open AI policy” for both homework and exams.

“We need to ensure our students become strong, contributing citizens and leaders—not just for the next year but for the next 40 years,” Ranade emphasized.