When students in communications professor Michael Schmierbach’s research methods course copied and pasted the prompt for their assignment into ChatGPT, they unknowingly fell for a hidden trick. The usage of small, white text embedded in Canvas assignments helped Schmierbach reveal their use of artificial intelligence (AI) technology.

The media studies professor specializes in news credibility, video game difficulty, perceived media influence, social media, and news. Over the years, he became more aware of the rise in AI technology and how it could be used by students.

After discovering several apparent AI-generated responses, Schmierbach devised a detection method. He created an assignment and embedded small white text onto Canvas. When students copied and pasted the assignment into AI bots, the bots detected the white text and answered accordingly.

“I thought, well, AI just does what you tell it. So if the students are not actually sitting there and thinking and crafting the query themselves but are just copying the query, then why not have them copy over some additional parts of the query?” said Schmierbach.

Schmierbach didn’t conduct this experiment to outsmart students using AI but rather to explain his concerns surrounding students’ lack of effort.

“It’s less intended to figure out the most sort of clever, determined cheats and more as just a first line of, well, if I’ve got that and I sit down and talk to you about why you need to actually be at least thinking about your own worry, it’s a pretty easy starting point of, I know you didn’t even try to think about the answer here,” Schmierbach noted.

The results of Schmierbach’s method varied, with several students clearly submitting AI-generated responses. He noted that many students who do not use AI technologies and maintain academic integrity might have been frustrated by their classmates’ lack of effort.

When thinking about education in the bigger picture, Schmierbach noted that AI isn’t always accurate and shouldn’t be used to replace learning but as a tool for editing and summarization. He believes that AI can’t replace the quality of work humans can produce.

“We’re not assigning these assignments because the results are the kinds of things you’re going to do in your professional life. We’re assigning them because these are students at sort of an early or middle point in their academic progression, and they’re trying to learn the skills to get to where they can do the actual professional work,” said Schmierbach.

Schmierbach isn’t the only professor who has noted the growing impact of AI technologies on academic integrity and education. Mila Sanina, an experienced journalist with over 15 years of experience, has expressed that AI technologies have caused her to redefine her role as an educator.

“It has forced me to consider what is my role in making sure that academic integrity is still the value that we hold paramount. That is the most important thing because it is influencing the value of a Penn State degree,” she said.

Sanina feels that AI cannot replace education and learning. She has tried to cover responsible uses of AI technologies in her journalism classes and thinks universities need to have deeper discussions on AI’s role in education.

She also expressed concerns surrounding AI detection methods, noting cases in which students whose first language isn’t English have had their work falsely cited as AI. Sanina feels that ethical considerations involving transparency and critical thinking should be made when dealing with AI technologies.

“I think there should be questions on, “Why are you using it?” What is the value of education? Is it an aspect of education of the 21st century, or a quarter of the 21st century is to teach you how to think better than AI?” Sanina said.

On the other hand, computer engineering professor Kyusun Choi has taught at Penn State for over 30 years and finds that AI technologies have revolutionized his teaching methods.

“Everybody uses calculators, so to me, the AI tool is just like a calculator. It helps people to quickly search and come up with the answers,” Choi shared.

The computer engineering professor understands that students must comprehend how to synthesize and analyze independently. However, he asks that his students use AI to its fullest potential for efficiency.

When teaching students how to design a computer chip, he noted that AI allowed him to simulate which design was better regarding speed and power. Although he favors AI technology and its usage in education, he also understands that it can be used for academic dishonesty.

“AI is a tool, and we have not fundamentally changed anything, but we did definitely change the speed of the computing,” noted Choi. “It’s just a tool, and to me, similar to the calculator long ago, and it is a tool that helps everyday activities and work be accomplished quicker and in a more efficient way.”

With growing concerns surrounding AI’s use in education, many cannot imagine a time when AI wasn’t easily accessible to the public. Former Penn State chemistry professor and advisor Joe Keiser, an educator before the time of AI, provided his own input.

Keiser taught at Penn State for 29 years and was the former director of General Chemistry Laboratories. He mentioned that academic dishonesty has always been a factor in education. To reduce cheating, he implemented in-person lab reports at Penn State.

“We used Turnitin, which helped, but students still cheated. When they did cheat, it was a huge mess. You would have to have meetings, and students would fight you, and you had to spend a lot of time digging out the bottom of what was going on,” Keiser said.

Keiser believes that AI technology has both benefits and drawbacks. During meetings with the undergraduate chemistry advisory group, he noted that although students can use AI to cheat, many professors have found it a positive educational tool.

“Something else that came up in this undergrad advisory group, which really impressed me, was how you can have AI do grading and tutoring that you would have normally had to have a person do,” Keiser shared.

He noted that one undergraduate professor shared how he uses AI technology to tutor students and advance their comprehension.

“I think there’s opportunities, plenty of them, to use AI to cheat, but I think there’s going to be also many opportunities to make the learning experience better for students,” said Keiser.

With various perspectives surrounding the role of AI in education and its influence on academic integrity, one question remains: Should AI be embraced or criticized in the classroom?