Recent research published in journals Telematics and informatics This suggests that college students are increasingly relying on artificial intelligence tools due to a strong fear of falling behind their classmates. This dependence tends to create a psychological tug-of-war, resulting in severe anxiety, along with a deep fear of being judged for using the technology when it is not available. These findings provide evidence that universities need to establish specific guidelines to help students avoid the mental stress of using artificial intelligence in their classes.
The integration of artificial intelligence into higher education is progressing rapidly. Today, 80-90% of college students use generative artificial intelligence tools like ChatGPT to draft essays, summarize what they read, and find information. Generative artificial intelligence refers to computer systems that can create new text, images, or code based on patterns learned from vast amounts of data.
Despite this widespread adoption, many universities have yet to establish consistent rules for how to use these tools. This lack of policy leaves students guessing about where the line lies between acceptable aid and academic dishonesty. Ambiguity can lead to stress as students try to understand what the teacher considers fair play.
“Since spring 2023, we have noticed a significant gap between student behavior and university policy,” said Chun-sik Lee, associate professor of communication at the University of North Florida and co-author of the study. “Given these gray areas, we wanted to investigate what causes students to become overly reliant on AI, and the associated negative psychological effects.”
Previous research on smartphones has documented a phenomenon called nomophobia, the anxiety people feel when separated from their mobile devices. The authors adapted this idea to modern technology and coined the term “noAIphobia” to describe the restlessness and anxiety that students experience when they are no longer able to access the digital assistants they normally use. NoAIphobia refers to a specific type of performance anxiety caused by the absence of tools that a person has come to rely on.
At the same time, the authors proposed that students may suffer from bias due to the use of artificial intelligence. This stigma includes fear that relying on computer programs will lead to negative evaluations by peers and mentors, such as being seen as lazy or unoriginal. The researchers aimed to map out how these various motivations and fears interact to influence whether students continue to use technology.
To investigate these psychological dynamics, researchers recruited 393 U.S. college students through an online survey platform. To participate, students must be currently enrolled in a college or university and have used generative artificial intelligence tools in their academic work this semester. The average age of the participants was approximately 29 years old, and the participants were approximately evenly divided between undergraduate and graduate students. Approximately 54 percent of participants were women.
Participants completed a detailed questionnaire designed to measure motivation and psychological state. They rated a series of statements on a numerical scale to indicate the extent to which they agreed or disagreed with various concepts. For example, the study measured efficiency expectations, the idea that technology saves time and effort. We also measured quality expectations, or the belief that the tool would improve student performance and overall quality of work.
Another concept measured in the study was competitive fit. This term refers to a student’s tendency to behave in the same way as their classmates simply to avoid being at a competitive disadvantage. The researchers also used specific questionnaires to measure students’ reliance on artificial intelligence, feelings of AI phobia, concerns about academic integrity, and fear of social bias. Finally, the survey asked how likely students were to continue using these tools in the future.
Scientists used advanced statistical modeling to analyze the survey responses and map the relationships between all these different factors. The analysis revealed that fear of falling behind colleagues is the strongest predictor of reliance on artificial intelligence. Competitive fit was more important than practical motives such as wanting to save time or produce better work. When students believed that their classmates were gaining an academic advantage through artificial intelligence, they felt pressured to rely on the technology themselves to avoid being put at a disadvantage.
The researchers also found that expecting a tool to save time and effort was a stronger predictor of dependence than expecting the tool to actually improve the quality of work. This suggests that many students view these programs primarily as shortcuts rather than learning aids. Regardless of the initial motivation, high levels of dependence were associated with two opposing psychological states.
“The biggest lesson is that students’ AI addiction is not just caused by laziness, but also by socialization and anxiety,” Lee told PsyPost. “We found that students are in a psychological dilemma. On the one hand, they experience ‘AI phobia,’ or the fear that they won’t be able to complete their work if they lose access to the AI. On the other hand, they face ‘AI stigma,’ or social concerns that others will view them as lazy or cheating.”
Students with a higher reliance on artificial intelligence were more likely to experience significant non-AI phobia. The thought of completing school assignments without a digital assistant made them nervous and anxious. This deprivation anxiety was strongly associated with the desire to continue using the tool. Students essentially wanted to maintain access to avoid the psychological discomfort of working without a program.
On the other hand, dependence also predicted higher levels of AI use of stigma. The more students relied on technology, the more they worried that their peers and professors would label them fraudsters. This fear of social evaluation was associated with a decreased motivation to continue using the tool and caused psychological conflict in students.
“We believe that the long-term de-skilling effect of AI could manifest as a resolution of AI phobia,” Lee added. “One of the interesting findings from this study is that AI phobia is positively correlated with students’ intention to continue using AI. At the same time, social concerns about AI use manifest as prejudice against AI use, which is negatively correlated with students’ intention to continue using AI.”
The authors noted that students’ personal concerns about academic integrity acted as glasses that magnified this stigma. For students who were deeply concerned about educational ethics and cheating rules, the relationship between technology dependence and feelings of shame was significantly stronger. These highly ethical students felt the burden of prejudice more strongly than students who cared less about academic rules.
“This study suggests that students are not simply taking the easy route of using AI,” Lee explained. “Rather, it reflects a real psychological dilemma that many college students experience.”
Although this study provides useful information about modern educational technology, it also has some limitations. This study is based on a single survey conducted at a specific point in time. Because of this design, researchers cannot prove strict causal relationships between variables.
“This is a one-time study, not a long-term study or experimental design,” Lee pointed out. “Thus, the reader should take care in interpreting the direction of causality.” Psychological states and technology dependence may interact in a continuous loop rather than operating in a linear manner.
The student sample was also collected through a specific online survey panel, so the results may not be completely representative of all college students in the country or around the world. Future studies could follow students over months or years to see how these dependencies and anxieties develop over time. Tracking behavior longitudinally could help scientists understand whether these feelings of bias eventually fade as technology becomes more normalized in society.
Future research could expand beyond mere academic challenges to see if these same psychological patterns hold true when people use artificial intelligence for personal entertainment or creative hobbies. Investigating whether this dependence ultimately reduces students’ actual critical thinking and problem-solving skills would also provide a more complete picture of technology’s long-term effects. To fully understand a technology, it is necessary to examine both its cognitive and emotional costs.
In the meantime, the researchers suggest that universities need to create specific and transparent policies regarding acceptable uses of artificial intelligence. Establishing clear boundaries can help reduce the bias and ethical confusion that students currently face. Providing this structural support may help minimize the psychological distress associated with these modern academic tools.
The study, “Motivational factors and psychological correlates of AI dependence in college students: Competitive conformity, AI phobia, and AI use bias” was authored by Junga Kim, Chunsik Lee, and Joon Soo Lim.
