Artificial intelligence is something that has grown rapidly in the past few years. With this powerful tool on its exponential rise, schools have had to find ways to combat unwarranted use of AI plagiarism by students. At the forefront of fighting back against generated answers has risen a new problem for students; AI detectors implemented by teachers are flagging human writing as forged by the technological hand. 

North Penn, among other schools, has utilized programs that allow teachers to quickly scan student submissions of work for AI usage. However, it isn’t always accurate. Micheal Botti, a North Penn technology support specialist, explains the possibility of mistakes made in originality reports on these sites.

“We have Turnitin, but it’s not recommended that teachers use it to detect AI, because it has a 2-4%, false positive. So, when you’re talking about a school this large- 3,000 students- it’s anywhere from 60 to 120 students that could be falsely accused of using AI,” Botti explained. 

Students across the school have faced issues with teachers where they were falsely accused of plagiarizing from AI. Hayley Moylan, a North Penn junior, was one of them, and her grades took a hit from it. 

“In my situation, it was a reading journal that was flagged for plagiarism. [My teacher] gave me a zero on the assignment despite the fact that the history showed me writing it. In the end, I was not able to make her believe me and I took the 0% on the assignment, which dropped my grade for the marking period to a C,” explained Moylan.

With AI usage becoming so prevalent and easily accessible by all, it’s something that will, without a doubt, show up in cases of plagiarism. However, that doesn’t mean that other students should be mistrusted when their work was written the right way. 

“I felt that the situation as a whole was very unfair, considering that there are endless platforms for people to express their ideas and analyses on the internet. Because of that, the probability that someone else posted a similar analysis to mine is very great. I’ll admit that this is a tricky situation, but I do feel that teachers should have more trust in their students,” Moylan said. 

Another concern is that neurodivergent students and English language learners are at risk for getting falsely flagged, because their writing style in some cases can look similar to what AI might generate. 

Trust in students is just the starting point for schools to overcome the issue of false AI flagging. Botti shared that various technology leaders advocate against using originality report platforms at this point. Instead, he recommends that teachers put in the work to scan assignments themselves by getting to know what kind of writing their students are bringing to the table.

“Teachers should be getting to know their students’ writing. It’s also dependent on the grade level. Sometimes it’s really evident when a student is using a certain vocabulary or sentence structure that can usually tip off a teacher to think if the student used AI,” Botti said. 

Botti also encourages teachers to consider why their students might be using AI in the first place. Taking a look to gauge student interest and understanding can prevent plagiarism from happening in the first place. Considering all of this, individual understanding of each student’s skill set is something that teachers should be striving to achieve if they want to prevent plagiarism from AI. 

Furthermore, rather than casting aside AI, it might be useful for teachers to gain familiarity with it and push for it in the classrooms even more. In today’s world, AI is near impossible to avoid; therefore, it should be embraced.

“You’re going into the real world where it exists, and you’re going to have to enter a job market at some point that you’ll need to be able to leverage AI to help you complete certain tasks. My philosophy so far has been encouraging teachers to use it so that they understand how it works and know some of the drawbacks and limitations, so that we can figure out how we empower students to be able to use it,” Botti said.

“I would hope that teachers find a way- and I know it’s scary- to make sure that students have the writing skills and behind them to be able to leverage AI. We need to look for ways to teach students how to use it and balance their use with their authentic thoughts and authentic ability to write because they’re going to need that skill eventually,” Botti continued.

Academic integrity is a standard that should be upheld without wrongly working against students who put forth their own work. Especially as artificial intelligence grows to learn to evade detection as it gets more advanced, using scanners such as Turnitin aren’t always going to be accurate. Artificial intelligence isn’t going away anytime soon; in fact, its prevalence can be seen increasing more and more. Therefore, it has to be handled delicately and navigated with methods that help students, not harm them. 

“It’s here, the genie is out of the bottle now, we have to figure out how we best move forward,” said Botti.