Key facts
- UC Berkeley math professor Zvezdelina Stankova admitted to using AI to edit an op-ed.
- The op-ed criticized a "severe" math deficiency among students.
- AI detection software indicated a significant portion of the op-ed was AI-generated.
- Stankova stated the article was the result of extensive human work, with AI used for research and editing.
- The incident has ignited a broader discussion about AI use in academia.
A mathematics professor at the University of California, Berkeley, Zvezdelina Stankova, has admitted to using artificial intelligence to assist in editing an op-ed she wrote for the San Francisco Standard. The article, published last week, criticized a "severe" math deficiency among students, suggesting they were "five to eight years" behind in fundamental concepts.
Journalists from Berkeley's student newspaper, the Daily Californian, flagged the op-ed after AI-detection software indicated that 33% of the text may have been generated or assisted by AI. In response, Stankova confirmed she used AI software for editing and to locate relevant documents, but emphasized that the analysis and the article itself were the result of "several hundred person-hours of intensive human work," with approximately 80 hours being her own.
The San Francisco Standard stated that while AI may assist, their expectation is that humans are responsible for all published content. Their understanding was that Stankova's piece reflected her and her colleagues' original analysis and expertise.
Stankova's admission has ignited a debate on social media about the ethical use of AI by academics. Critics have pointed to the irony of using AI in an article discussing student unpreparedness, while others defend the professor, arguing the AI usage is a distraction from the central argument about reinstating standardized tests in college admissions.
The debate over standardized testing in UC admissions has intensified since 2020, when the UC regents voted to phase out the SAT and ACT as requirements. Several peer institutions that initially followed suit have since reinstated testing requirements. Stankova's op-ed cited data suggesting a tripling of students with "severe deficits" in calculus readiness after test-blind admissions were introduced. She also noted that AI was used in editing a June letter signed by over 3,000 UC faculty members supporting admissions testing.
The UC academic senate is currently reviewing the potential reinstatement of standardized tests, with any changes possibly affecting admissions for fall 2028. The UC system also has an AI council, though a member noted that AI detection tools can be unreliable, a sentiment echoed by Stankova who called the AI usage debate "orthogonal to and a distraction from" the initiative to reinstate standardized tests.