Psychosocial and Academic Correlates of Generative AI Use and Its Perceived Influence among College Students

In the quiet hum of a college dorm room or the busy chaos of a library, a new kind of companion has quietly taken up residence. It is not a person, but a form of artificial intelligence capable of generating text, ideas, and advice on demand. For many young adults, this technology has become a standard tool for navigating schoolwork, helping to explain complex concepts or draft essays. Yet, its role is expanding beyond the classroom. Students are increasingly turning to these digital systems for personal reasons, seeking comfort when lonely, advice on difficult relationships, or help managing anxiety. This shift raises a fundamental question for researchers: who is using these tools, and why? The answer likely depends on a student's internal world—their stress levels, their mental health, and how well they are adjusting to life at university. Understanding these patterns is crucial because the technology is not a neutral mirror; it reflects and responds to the specific needs and vulnerabilities of the person holding the device.

A team of researchers at Michigan State University set out to map this landscape by listening to 135 undergraduate students. They wanted to know not just how often students used generative artificial intelligence, but what drove them to it. The study revealed a clear divide in how students interact with the technology. Nearly all of the students, about 94 percent, used the tool for academic learning. It was a universal part of their educational toolkit. However, the picture changed when looking at personal life. About 45 percent of the students reported using the technology for at least one socioemotional purpose, such as seeking relationship advice, finding companionship, or getting mental health support. This suggests that for almost half of the student body, the technology has stepped into roles traditionally filled by friends, counselors, or family members.

The researchers found that a student's emotional and academic state strongly predicted how they used the technology. Students who reported lower stress levels and higher overall well-being tended to use the tool more frequently for learning. These students also viewed the technology as having a positive influence on their ability to solve problems. In contrast, students who were struggling with mental health challenges, feeling lonely, or lacking support from friends and family were the ones turning to the technology for emotional reasons. Those with more mental health problems used the tool more often for advice on coping or managing anxiety. Similarly, students with less social support were more likely to use it for companionship or to chat when feeling isolated.

This pattern extended to how students felt about the technology's impact on their lives. Students who were doing well academically and emotionally saw the tool as helpful for problem-solving. But for those facing difficulties, the perceived benefits shifted toward the personal realm. Students with higher stress levels felt the technology had a more positive influence on their social relationships. Those with mental health struggles or low social support felt it helped them understand themselves better. The data also showed a direct link between usage and perception: the more frequently a student used the technology for emotional support, the more helpful they rated it compared to other sources like family, friends, or professionals. Among those who used it for emotional support at least once a week, a significant majority felt it was at least as helpful as, or even better than, other sources of support.

Despite these findings, the researchers are careful not to declare the technology a perfect solution. The study was conducted at a single university with a specific demographic, and the design was a snapshot in time, meaning it cannot prove that the technology causes changes in a student's well-being or that a student's well-being causes them to use the technology. It is possible that the tool helps, or that it simply fills a gap left by other services. The authors note that while the technology offers immediate, private, and seemingly nonjudgmental responses, this very appeal could lead to overreliance. A student might find comfort in a digital conversation, but that does not guarantee a lasting improvement in their mental health or relationships. The study suggests that for students with unmet psychological needs, this technology is becoming a primary resource, but it remains unclear whether it is a helpful supplement or a substitute for the human connection they truly need. The findings serve as a reminder that the impact of artificial intelligence is not uniform; it is deeply shaped by the individual circumstances of the person using it.