Published October 27, 2026
A new study has revealed one of the most alarming effects of using artificial intelligence (AI) for problem solving. Researchers found that it only takes 10 minutes of any AI tool usage to erode motivation and the people who use AI more often are likely to give up easily when facing difficult tasks.
The research conducted by scientists from Carnegie Mellon University, MIT, Oxford, UCLA and the University of California, Berkeley, found that relying on AI tools for quick answers may come at the cost of independent thinking.
The study, published at the 2026 Conference on Language Modeling, involved 1,222 participants who were asked to solve mathematical problems and reading comprehension questions.
The participants were divided into two groups: One was allowed to take help from AI while the other group had to complete tasks on their own.
At first, the group which was allowed to use AI performed better and skipped less questions compared to the other group; however, when researchers withdrew access to AI chatbots, the participants who were previously using AI struggled to solve similar questions and were more likely to give up.
The findings were consistent across three experiments involving 354, 667 and 201 participants, respectively.
Researchers found that participants who relied on AI to provide direct answers experienced a greater decline in persistence than those who used it for guidance.
Grace Liu, the study's lead author and a doctoral student at Carnegie Mellon University, explained that struggling with difficult problems plays an important role in developing new skills.
She warned that AI tools offering immediate solutions could deprive users of the mental effort needed to strengthen their independent capabilities.
Researchers believe that people who become accustomed to receiving answers within seconds may start considering traditional problem-solving methods unnecessarily time-consuming.
However, the study examined only short-term effects and did not establish whether using AI for brief periods causes permanent damage to cognitive abilities.
The scientists suggested that AI chatbots should function more like tutors, helping users reach answers themselves rather than simply providing ready-made solutions.