A student who said she submits AI-generated book summaries in place of reading the assignments and writing her own summaries poses for a portrait at San Francisco State University last month. She was one of a dozen students at SFSU and at UC Berkeley who spoke freely about how they use AI to do some or all of their schoolwork.
Manuel Orbegozo/For the S.F. Chronicle
A UC Berkeley senior sat eating a sausage sandwich near Sather Gate the other day when a reporter asked if he had ever used AI to write a paper instead of writing it himself.
Yes, he had. “But not verbatim. I always rewrite it. I always like to add my own pizzazz,” he said, after being assured of anonymity to encourage candor about the sensitive subject of cheating in the age of artificial intelligence.
University of California and California State University campuses, like others around the country, all outlaw cheating — commonly understood as “the attempt to pass off work as your own without having done the work,” said Irina Raicu, director of internet ethics at Santa Clara University’s Markkula Center for Applied Ethics.
Article continues below this ad
But in just a few years, generative artificial intelligence chatbots like ChatGPT, Google Gemini and Claude have become part of students’ daily life. With AI often closer at hand than a ballpoint pen — able to write brainy essays, answer any question and summarize entire books for students who never crack them open — what does cheating mean to them any more?
UC and CSU are still figuring out how AI fits into their educational mission: how to embrace and encourage the technology, while discouraging its abuse, as busy students are tempted to outsource their schoolwork to chatbots they believe can do it better and faster than they can.
San Francisco Chronicle Logo
Make us a Preferred Source to get more of our news when you search.
Add Preferred Source
Meanwhile, guidance from UC Berkeley’s Academic Senate and CSU acknowledges that AI has created potential gray areas in ethical decision-making. They emphasize that instructors are expected to use their syllabi to spell out acceptable uses of generative AI tools for their courses, while reminding students they are still required to produce their own work.
But not all students appear to be checking the rules. Conversations with a dozen of them at UC Berkeley and San Francisco State yielded story after story about how they parcel out their reading and writing to machines, leaving open the question of what they are actually learning.
Article continues below this ad
“I have an A because I use Chat,” a San Francisco State psychology major said of her statistics grade. “If there’s a math problem, instead of me trying to study for a class I don’t really care about, Chat will just solve the problem for me.”
‘Everybody’ uses it, ‘since high school’
While AI mimics intelligent behavior, it so far mimics ethical behavior only with specific prompting.
And that is up to students.
One pre-med student at UC Berkeley, for example, said that before asking AI to help her get started on, say, a physics problem, she must warn it: “Don’t tell me the answer!”
Article continues below this ad
Two of the 12 students interviewed at UC Berkeley and San Francisco State said they abstained from AI altogether, citing the technology’s excessive use of water and energy in one case, and an unwillingness to “lose my voice” in the other. Four students said they used it only for legitimate, tutoring purposes — to explain math and science complexities or to quiz themselves — and never to do their work for them.
But six said they use AI to do some or most of their work, with some insisting that this isn’t cheating. Others said that it is. Some criticized professors for prohibiting AI, while others claimed their instructors don’t actually want them to do the assigned reading and are fine with AI-generated summaries.
The Chronicle did not conduct a scientific survey but listened as students discussed the tools that make their academic experience unlike that of any in the history of the planet.
“I know it’s cheating, and I do feel a little oof about it,” a San Francisco State freshman said as she bit into a burrito in the student union.
She types her notes into ChatGPT then asks it to turn them into polished assignments for English and communications classes.
Article continues below this ad
“It’s basically what I submit. I’ll say, this sounds a little robot-y, so I’ll tweak it a little,” she said. “But it’s getting my stuff done, so I don’t feel horrible about it.”
At UC Berkeley, a junior majoring in media studies said she relies on AI for all assignments.
“The heavy lifting I use AI for is my reading,” she said, naming a site that turns reading material into short podcasts that she listens to before class. “Kind of like, abridged.”
What about using AI for her writing? “Yeah, 10 out of 10” assignments, she said. “Though I try not to plagiarize off ChatGPT.”
Her definition of “plagiarize,” at least, was hers alone.
Article continues below this ad
She’ll have a “big sploosh of ideas,” which she types into Chat. “I’ll be like, can you condense my ideas into a few sentences? And it will. I’ll usually copy what it says. I’ll consider it mine because they’re my ideas,” she said.
At San Francisco State, a student majoring in speech and language said she couldn’t remember the last book she had read for school, though she had recently been assigned at least three.
“We weren’t supposed to read the books,” the student claimed. “Just summarize them.”
To achieve this, she used two AI sites: one to summarize, and the second to ask if the summaries were good enough.
The second site told her the summary of “The Glass Castle,” a memoir, needed more work.
“For that one, I did not really understand,” the student said. So she returned to the first site and told it to do a better job.
“Everybody uses” it, she said. “Since high school. It’s like a tool.”
Colleges race to adapt
The existence of intelligent tools doesn’t change the definition of ethical, fair behavior, said Ahmet Palazoglu, chair of UC’s Academic Senate.
“The core principle that students must do their own intellectual work has not changed,” he said.
All UC and CSU campuses have academic integrity policies that explicitly require students to submit their own work and to attribute material they didn’t create to the source that provided it.
In the UC system, the Academic Senate’s committee on educational policy provides guidance on standards for academic conduct, while faculty at each campus decide how to put the ideas into practice.
UC Berkeley’s Academic Integrity policy, for example, requires “proper attribution” in all work, including “drafts and homework assignments!” It says cheating “includes fraud, deceit, or dishonesty in an academic assignment” or using materials that are “prohibited or inappropriate in the context of the academic assignment in question,” such as when a syllabus prohibits the use of AI.
The school’s Code of Student Conduct warns that “cheating, fabrication, plagiarism and facilitating academic dishonesty” are subject to discipline.
Traditionally, violations were considered rare. Then, as the rise of the internet enabled easy copy-paste plagiarism, cheating detection tools became common. Then, “contract cheating” arose, as students bought custom-written papers from human fraudsters in the pre-Chat era.
But AI now facilitates cheating that is so easy to produce and hard to detect that experts fear an epidemic of cheating is unfolding.
This new reality is having a domino effect at UC Berkeley, where campus officials have posted a flurry of wisdom and resources to help professors minimize cheating. There are workshops and no fewer than six groups of big thinkers discussing how to co-exist with such tools, while UC’s systemwide Academic Senate and Office of the President have four others doing the same.
“We take the challenges seriously,” said Palazoglu of the Academic Senate. “The answer is not to pretend AI does not exist, but to be clear about what constitutes real learning. To redesign learning assessments where needed, and to have frank conversations with students about why doing the work themselves matters.”
The university must “help students see that abuse of AI’s capabilities simply undermines their ability to take full advantage of a once-in-a-lifetime opportunity,” said Oliver O’Reilly, UC Berkeley’s vice provost for undergraduate education. “When students cheat, they only cheat themselves.”
Yet where persuasion fails, guardrails are needed, O’Reilly said. “Instructors are adapting by changing assessments, insisting on proctored exams and redesigning courses.”
CSU’s approach: Embrace the tools, accept the risks
CSU declared itself the “nation’s first and largest AI-empowered university system” last year when it announced deals with ChatGPT’s inventor, OpenAI, and other tech companies to expand students’ access to AI tools.
“At the CSU, we have two imperatives: to equip our students with the skills to leverage these powerful tools, and to transform our own institutional practices through AI to better serve the largest public university system in the nation,” Ed Clark, the university’s chief information officer said at the time.
A year later, on April 1, CSU released “Ahead of the Curve,” a report on AI use that surveyed 80,626 students and 13,434 faculty and staff across its 22 campuses.
Nearly 80% of the students said they would not submit AI-generated work as their own. That left more than 16,000 students, about 20%, who did not check that box.
“Most students are not using AI to cheat,” said Jason Maymon, a CSU spokesperson. “The idea that AI is driving a widespread increase in academic dishonesty at the CSU or in higher education is misleading.”
Yet thousands of student responses left open the possibility that they do submit AI-generated work as their own.
Asked to comment on the students who confessed to doing just that, Maymon accused the paper of “promoting a narrative that students simply want to use AI to cheat.”
Focusing on dishonesty, he said, “threatens to undermine the responsible, ethical ways students are already engaging with these tools at the CSU.”
CSU’s student conduct policy prohibits “forms of academic dishonesty that are intended to gain unfair academic advantage” — and a webpage, “Ethical and Responsible Use of AI for Students,” that gives classroom examples.
Preventing cheating has never been easy. It’s harder now.
“We cannot punish students for using something they use innocently on a daily basis,” said Carlos Montemayor, chair of the San Francisco State philosophy department who writes widely about AI.
At the same time, he said, “it’s very important that students understand that they should not present ideas as if they were their own when they were the result of someone else’s creativity.”
What instructors see: ‘It’s very obvious’
Many professors and instructors say they aren’t giving up, even if all they can do is spell out expectations on their syllabi, change their teaching methods and cross their fingers.
Violations abound, which drives at least one UC Berkeley graduate student instructor nuts. The social scientist grades many papers and requested anonymity to speak freely about her students.
“It’s like reading the same essay over and over with the same verbiage,” she said, her voice edged with frustration. “It’s very obvious” they used AI.
She suspects that at least half of her students are not doing the work themselves. She is certain that 20% aren’t. She gives them an F and asks to see the assignment’s version history.
“They’re very obstinate and defensive” when confronted, she said. They exhibit “moral grandstanding” and say, “I would never use AI! But looking at their version history, it’s very obvious. They never cop to it.”
The media studies major shrugged. Using AI to do school work “is OK because everyone has so many things to do,” she said. The fault lies with professors who are “old school” and “afraid of change.”
“If you don’t use the new tools you are given, you’ll be left behind,” she said. “This is what happened to Blockbuster,” the defunct video chain.
‘You’re cheating yourself’
One emerging area of research is whether the human brain suffers by having AI do its thinking, the way muscles atrophy from lack of use. It’s called cognitive debt, and scientists say they can measure it.
One preliminary study from the MIT Media Lab last June, for example, found that students’ brain power was reduced when AI wrote for them, raising “concerns about the long-term educational implications” of relying on machines to do our cognitive calisthenics.
To Raicu of the Markkula ethics center, “the danger is that we have a bunch of people who get ‘faux educated.’ People who have degrees but who don’t really know how to think and how to write.”
One San Francisco State senior came to realize that on his own.
In 2022, he said, he was a community college student working at a restaurant “30 terrible hours” a week. He found English 101 especially tough and began copying book reviews from the internet into an AI tool.
“It paraphrased everything I pasted in,” he said, then re-paraphrased it until “it was close to what I would have written.”
He passed. Eventually, he took a psychology class he loved and began working at an after-school program.
“It just really changed me. I saw the impact I had for those kids,” he said. “I realized I wanted to be a school counselor.”
He said he no longer uses AI to do his work. “You’re cheating yourself out of the value of your own education. I’m aware of that.”
But without it, he said, “I would have failed. I might have dropped out and gone into a trade. I am very, very grateful” for that technology.