Generative artificial intelligence is transforming how people learn, work, and solve problems, but researchers are increasingly warning that relying too heavily on these tools could weaken essential human thinking skills over time.
Experts say abilities such as memory, decision-making, critical thinking, and problem-solving may be at risk as AI becomes a routine part of daily life. While AI can boost productivity and provide quick answers, studies suggest that frequent dependence on these systems may reduce people’s willingness to think independently.
A recent American-British study involving 1,222 participants found that AI tools helped users perform better on arithmetic and reading comprehension tasks in the short term. However, researchers observed that participants who regularly relied on AI performed worse when the tools were removed and showed less persistence in solving problems on their own. The study is currently undergoing peer review.
Researchers noted that persistence is a key factor in long-term learning and skill development. According to Grace Liu, a doctoral researcher at Carnegie Mellon University and the study’s lead author, AI can unintentionally reduce valuable learning opportunities by providing instant solutions instead of encouraging users to work through challenges themselves.
Unlike previous technologies such as calculators, which assisted with calculations while leaving the reasoning process to users, generative AI can perform a wide range of intellectual tasks, making it far more influential in shaping human thinking habits.
The findings align with other research, including a widely discussed 2025 MIT study that reported students using generative AI to write essays demonstrated lower levels of critical thinking. Researchers describe this growing dependence as “cognitive offloading” or even “cognitive surrender,” where people increasingly shift mental effort to AI systems.
French cognitive psychologist Johann Chevalere explained that the human brain naturally seeks to conserve energy, often choosing the easiest path to complete tasks. He warned that if people consistently allow AI to perform mentally demanding activities, the brain may gradually weaken the neural connections associated with those skills.
In response to these concerns, major AI developers are introducing features designed to promote active learning rather than simply delivering answers. OpenAI’s ChatGPT now includes a Study Mode that guides users with questions and hints, while Google’s Gemini offers a Guided Learning feature aimed at encouraging deeper understanding. Microsoft has also added reminders in Copilot urging users to verify AI-generated information and remain actively engaged in evaluating responses.
Despite growing evidence, scientists say more long-term, large-scale research is needed to fully understand how generative AI will affect human cognition. Until clearer answers emerge, experts recommend using AI as a tool to support learning rather than replace independent thinking, emphasizing that balanced use will be essential as society adapts to this rapidly evolving technology.
