Pedagogical readiness in a structural equation model of generative artificial intelligence use and cognitive offloading among preservice teachers

In the modern classroom, a new kind of helper has arrived, one that can write essays, solve problems, and explain complex ideas in seconds. This technology, known as generative artificial intelligence, has changed how students approach their schoolwork. For many, it acts as a powerful shortcut, allowing them to finish tasks faster or understand difficult concepts with less struggle. However, this convenience comes with a catch. When a student relies too heavily on a tool to do the thinking for them, they risk skipping the mental effort required to truly learn. This phenomenon is called cognitive offloading: the act of shifting the mental work of a task onto an external device. While this can be a smart strategy for managing heavy workloads, it becomes problematic if it replaces the deep thinking necessary for mastering a subject.

This issue is especially critical for people training to become teachers. These future educators are not just students using the tools for themselves; they are the ones who will eventually decide how to teach these tools to their own pupils. They face a dual challenge: they must learn to use the technology effectively for their own studies while simultaneously figuring out how to guide their future students in using it responsibly. The question for researchers is not just whether these future teachers are using the technology, but why they are using it, what rules they feel safe following, and whether their confidence in teaching with the tool changes how they rely on it for their own learning.

A team of researchers at Masaryk University in the Czech Republic set out to understand these dynamics among nearly seven hundred students training to be teachers. They wanted to see if the reasons people usually adopt new technology, such as how easy it is to use or how much pressure peers and professors apply, actually drive the use of artificial intelligence. They also wanted to know if clear rules from the university made students feel safer about admitting they used the tool, and whether feeling ready to teach with the technology actually stopped students from using it as a crutch for their own thinking.

The researchers gathered data from students across various stages of their teacher training, from bachelor's degrees to advanced master's programs. They asked detailed questions about how often the students used generative artificial intelligence, what tasks they used it for, and how they felt about the rules surrounding its use. They specifically looked at whether the students felt the university had clear guidelines and whether they felt safe admitting to using the tool without fear of punishment. They also measured the students' confidence in their ability to use the technology for teaching, a skill set known as pedagogical readiness, and compared it to how often they let the tool take over their own mental work.

The results revealed a story driven almost entirely by efficiency. The strongest reason students used the technology was simply that it made their work faster and more effective. The researchers found that the belief that the tool would help them get better results was the single most important factor predicting how often they used it. Surprisingly, other factors that usually influence technology adoption, such as how easy the tool was to operate, the pressure from friends, or even the influence of professors, did not significantly predict how often the students used the technology. The students were not using the tool because their professors told them to or because their friends were doing it; they were using it because it worked.

The study also uncovered a vital link between clear rules and student peace of mind. When students felt that the university's rules about using artificial intelligence were clear and consistent, they reported feeling much safer. This psychological safety meant they were more willing to be transparent about their use of the tool. However, the researchers found that this feeling of safety did not change how often the students used the technology. Whether a student felt the environment was supportive or risky, they still used the tool if they believed it was useful. The rules mattered for honesty, but not for the decision to use the tool in the first place.

Perhaps the most significant finding concerned the relationship between using the tool and the mental effort students put into their work. The data showed a strong connection: the more frequently a student used the artificial intelligence, the more they reported shifting the mental load of their tasks onto the machine. This suggests that as students use the tool more often, they are increasingly likely to let it handle the hard parts of thinking, such as formulating arguments or solving complex problems. This tendency to offload cognitive work was not reduced by the students' confidence in their ability to teach with the technology. Even those who felt very prepared and skilled at integrating the tool into a classroom setting were just as likely to rely on it for their own studies as those who felt less prepared.

This finding challenges the idea that simply knowing how to teach with technology will prevent students from becoming overly dependent on it. The researchers observed that while students felt confident in their personal ability to use the tool for their own studies, they felt much less confident about how to use it effectively in a teaching context. There was a large gap between their personal comfort with the technology and their professional readiness to guide others in its use. This suggests that being a skilled user does not automatically make someone a skilled teacher of that technology.

The study also highlighted that for the small group of students who did not use the tool at all, the reasons were not about technical difficulty or lack of access. Instead, they were driven by ethical concerns. Many of these students worried that using the tool would be a form of academic dishonesty or that it would weaken their own skills over time. Some even expressed concerns about the environmental impact of the technology. These students were not afraid of the rules; they were morally opposed to the practice.

Ultimately, the research paints a picture of a generation of future teachers who are pragmatic users of artificial intelligence. They are driven by the desire to be efficient and to get their work done well. While clear institutional rules help them feel safe enough to be honest about their usage, these rules do not stop them from using the tool when they need it. Furthermore, feeling ready to teach with the technology does not protect them from the temptation to let the tool do the heavy lifting for their own learning. The study suggests that teacher training programs need to go beyond just teaching how to use the technology. They must also help future educators understand the fine line between using a tool to support learning and using it to replace the thinking process entirely. Without this deeper understanding, even the most confident future teachers may find themselves relying on the technology in ways that undermine their own learning and, eventually, the learning of their students.