Excessive Dependence on AI May Increase Anxiety and Emotional Burnout Among Students

The use of artificial intelligence has become a common part of the educational process. But a new study shows: if AI turns into a replacement for one's own thinking rather than an assistant, it can negatively impact not only the quality of education but also the psychological well-being of students.

AI has become an everyday learning tool

Generative neural networks have already firmly entered the lives of universities. According to some estimates, over 90% of students use such services in one form or another: for information retrieval, preparing brief summaries, writing essays, or solving educational tasks.

While teachers and universities discuss the impact of such technologies on education, scientists are increasingly asking another question: is there a hidden dependency on AI among students, and what consequences can it have?

The study published in BMC Psychology indicates a worrying pattern. A team of researchers led by Wenlong Wang found that students who rely more on artificial intelligence tend to believe less in their own ability to handle educational tasks. At the same time, they often report higher levels of emotional burnout and anxiety.

How the study was conducted

The authors of the study surveyed 1,623 undergraduate students from China.

The participants filled out several psychological questionnaires that assessed:

• levels of academic stress;
• anxiety;
• emotional burnout;
• confidence in their ability to overcome difficulties and successfully complete assigned tasks.

Additionally, the researchers measured the degree of dependency on artificial intelligence. This assessed not the frequency of using neural networks, but how much the students themselves felt they relied on them for thinking, finding solutions, and performing academic work.

The researchers then analyzed the relationships among all the obtained measures.

Stress pushes towards using AI

The results showed a fairly clear trend. The higher the academic load and pressure, the more often students felt the need to rely on artificial intelligence tools.

However, along with the growth of such dependency, indicators of emotional burnout and anxiety also increased.

Data analysis showed that dependency on AI does not just accompany these states but may be a factor explaining their emergence.

Why self-confidence decreases

One of the key findings of the study was the influence of artificial intelligence on the sense of one's own competence.

Students who relied more on AI often felt that managing academic challenges by themselves was harder. This decline in confidence in their abilities was found to be associated with higher levels of anxiety and emotional exhaustion.

In other words, while artificial intelligence can ease the resolution of current tasks, it can simultaneously strip a person of the opportunity to gain their own experience in overcoming difficulties. And it is this type of experience that typically builds self-confidence and helps one cope better with stress.

The problem may be broader

The results align well with other studies from recent years. More scientific works indicate that regularly handing over intellectual work to artificial intelligence can gradually weaken the skills of independent analysis and reduce a person's confidence in their reasoning.

As complex tasks are increasingly solved by algorithms, the opportunities to train one's own thinking diminish.

What cannot be asserted yet

The authors emphasize that the study does not prove a causal relationship.

It is currently impossible to definitively say whether dependency on AI leads to decreased self-confidence and increased anxiety or whether the situation develops in the opposite direction—students who initially doubt their abilities start turning to neural networks more frequently.

To clarify this issue, long-term studies are needed that allow for observation of students over an extended period. Researchers also suggest testing whether educational techniques that train critical analysis of artificial intelligence responses help maintain independent thinking and reduce reliance on such tools.

AI should assist in learning, not learn instead of humans

The work of Wenlong Wang and his colleagues demonstrates that artificial intelligence is most beneficial when it serves as a support in the learning process rather than replacing it.

In practice, achieving such a balance is challenging. Moreover, the limitations of current generative models are well-known: they can make errors, provide inaccurate information, and even completely fabricate facts. That is why the use of AI requires not only convenient tools but also the preservation of the habit of analyzing information independently and making decisions.