Guilty Until Proven Human? AI Detection Raising Questions
By Gianna Difulio | October 28, 2026
When President's List student Gracie Giamartino opened her most recent graded assignment in her biology class at the University at Albany, the last thing she expected to see was an accusation of using artificial intelligence to write her paper.
The comment from her professor read, “Be careful with use of AI to help with writing formality,” and was accompanied by a grade of zero.
For Giamartino, who takes pride in her extensive vocabulary and grammatical skills, the feedback came as a surprise. She recognized that some of the characteristics in her writing, including em dashes and semi-colons, are also common in AI-generated text.
She immediately went to office hours to ask for an explanation.
The response, she said, was that her writing is overly formal. The way she structured her paper raised concerns because biology majors do not typically write in that format.
As generative artificial intelligence use becomes more common among college students, questions emerge about how professors can distinguish student's work from that of AI-generated.
A study by the University of California, Berkely's Center for Studies in Higher Education found that roughly 40% of students use AI at least monthly. However, this figure may not capture the full extent of student use if some are hesitant to disclose their use of technology.
“Honestly, I feel really bad for professors because I wouldn’t know what to do,” Giamartino said. “I don’t think AI detectors are very reliable because they themselves are AI.”
Giamartino's grade was changed after she had an honest conversation with her professor. She acknowledged that these situations can be difficult for instructors to navigate, particularly in larger classes when professors may not have the ability to familiarize themselves with students' writing habits.
Resolving a concern is not always as simple as having a conversation with your professor, especially when allegations move beyond the classroom.
“AI is tough because it is evolving on a second-by-second basis,” said Craig Brewer, UAlbany’s assistant dean of students for Community Standards.
The Office of Community Standards handles the final stage of the academic misconduct referral process.
Brewer said the office works closely with Academic Affairs, with alleged violations of academic integrity documented through the university’s violation reporting system, known as VAERS.
Brewer acknowledged that the process with suspected AI use is complex because of the newness of the technology, and its ability to retrieve more knowledge.
Through Community Standards, students have due process rights and opportunities to challenge and appeal allegations. Brewer said students can present evidence, witnesses and other relevant information to the appropriate chairs and deans of students.
Students are encouraged to exercise these rights and bring forward any information they may have.
“That’s ultimately what community standards wants to do; we want to get it right,” Brewer said. “We don’t want to punish people just to punish people.”
Students dissatisfied with a standards decision also have the opportunity to appeal to the decision. This allows students to challenge the outcome through every step of the procedures.
Giamartino's case was resolved within the classroom, but professors may refer other cases to university officials through formal reports.
UAlbany’s teaching and learning resources website provides guidance and resources for instructors responding to suspected AI use on student assignments. However, Brewer admits that faculty are still working on the best ways to address the challenges.
“I think that professors shouldn’t take the easy way out, and instead actually look at the writing from the student,” said Sergio Fuentes Rios, a sophomore student who had to rewrite an entire paper after being falsely accused of using AI.
As of Fall 2026, approaches to AI use and detection vary among UAlbany faculty and academic departments, leaving students to navigate these differences. Earlier this year, an Artificial Intelligence (AI) Taskforce was established at the university, and a proposed policy draft and guidelines were sent to students in September, with options available to leave feedback.
According to Brewer, the academic divisions of UAlbany are currently working on an artificial intelligence policy intended to apply clearer expectations for both professors and students, including guidance for AI use in the classroom.
These changes will hopefully be incorporated within the next couple of semesters.
He hopes the policy will create a sense of consistency for students and reduce confusion.
“We will never be perfect because AI has the ability to be more aware of everything. It just might take a little bit more time to get there [consistency],” Brewer said.
Comments