Artificial intelligence is increasingly shaping how people learn and process information, raising questions about its impact on cognitive skills and brain development. Recent research indicates that the widespread adoption of AI tools, such as chatbots, may be altering the way students engage with educational material, with potential implications for long-term learning.
A team from the University of California, Irvine, alongside researchers from the education company McGraw Hill, analyzed a decade of data from an online learning platform focused on mathematics, covering a period before and after the release of ChatGPT in 2022. The study examined millions of math problem sessions across a range of student ages, from fifth graders to college students. Researchers divided questions into two categories: text-based problems easily solvable by AI chatbots, and more complex visual or multi-step problems less accessible to such tools.
After 2022, students spent significantly less time per question on the text-based problems, and their immediate performance on these questions improved. However, when tested in controlled, supervised settings without AI assistance, student scores declined compared with pre-2022 levels. This suggests that while AI can accelerate problem-solving, it may also reduce deeper understanding and independent competency.
Additional studies highlight concerns that AI’s ease of access can lead to diminished persistence in problem-solving. One multi-institutional investigation presented evidence that adults accustomed to AI assistance tend to give up more quickly on challenging tasks they might otherwise solve, as the technology enables users to bypass the cognitive effort traditionally required in learning. Experts argue that bypassing the “productive struggle” involved in reasoning and analysis could undercut critical skill development and long-term retention.
Dr. Adam Green, a cognitive neuroscientist at Georgetown University, explains this phenomenon through the lens of the brain’s natural inclination to conserve energy, comparing mental exertion to calorie expenditure. He notes that younger learners, in particular, may miss opportunities to develop independent thinking if AI provides an easy alternative.
Nevertheless, not all findings are negative regarding AI’s role in education. A study conducted at Middlebury College involving over 200 undergraduates found that students who used AI tools to assist with essay writing produced longer and higher-quality work, and scored better on related multiple-choice tests. Importantly, when students later worked without AI support, those who had initially used AI as a supportive tool rather than a shortcut showed superior performance compared to their peers. This suggests that AI can enhance learning if integrated strategically to encourage active engagement rather than replace effort.
Overall, current research presents a nuanced picture: AI is reshaping how learners approach tasks, influencing both efficiency and cognitive engagement. The extent to which AI may impair or enhance learning appears to depend largely on how users employ these technologies—whether as shortcuts or as aids to reinforce understanding and skill development. Further study is needed to fully understand the long-term cognitive effects as AI tools become more deeply embedded in educational settings.