Research Suggests Overreliance on Artificial Intelligence May Undermine Human Cognitive Abilities

Artificial Intelligence Dependence Raises Concerns Over Long-Term Learning Impact

A recent study by researchers from MIT, Carnegie Mellon University, the University of Oxford, and UCLA has shed light on the potential consequences of relying heavily on artificial intelligence (AI) for learning and critical thinking. The research, which involved more than 1,200 participants, found that using AI chatbots for as little as 10 minutes can make people more likely to give up on difficult tasks and perform worse when the technology is removed.

The study, which focused on reading comprehension and maths problems, revealed that participants who relied on AI tools completed tasks faster in the moment but struggled more once the assistance was suddenly removed. Many participants gave up earlier or produced weaker answers compared to those who had never used AI support. Researchers described this trend as a possible "boiling frog" effect, where small and repeated dependence on AI slowly weakens human persistence and reasoning over time.

The study warned that while AI systems improve short-term productivity, they may also reduce the mental effort needed for long-term learning and skill development. In fact, the findings suggest that using AI mainly for direct answers can lead to a decline in performance, whereas using AI for hints or explanations did not have a significant impact.

Read also: OpenAI Board Member Sam Altman Reiterates Non-Involvement in Future Elections

The research highlights the need for balanced use of AI, especially among young users who are still developing core problem-solving skills. Education experts say that the study offers one of the clearest signs so far that convenience may come with a cognitive cost. The findings have not yet fully settled the debate on AI's impact on human intelligence, but they offer a crucial reminder that the long-term effects of AI dependence need to be carefully considered.

MIT assistant professor Michiel Bakker notes that the study should not be seen as a call to ban AI tools in classrooms or workplaces. Instead, he argues that AI systems should be designed to guide people toward solutions rather than simply providing direct answers. The study's findings arrive at a time when AI chatbots are becoming an integral part of everyday life for students, office workers, and businesses around the world.