The researchers recruited 350 Americans, who were asked to complete a set of tasks. However, only half were granted access to a chatbot for help

Think of the last time you navigated to a new part of the city without using GPS, or the last time you tried to call a friend whose number wasn't saved in your contacts. You can't remember, right? There was a time during the pre-smartphone era when our brains acted as vast filing cabinets of directions, dates, contacts, and more. But today, we've outsourced that to the devices in our pockets. Although that in itself should ring alarm bells for our future, the use of artificial intelligence appears to worsen this problem even further, as per an extensive study published on April 7.

With the use of AI becoming an invisible scaffold in our daily lives, researchers have provided the world with the first causal evidence that leaning on AI to assist with "reasoning-intensive" cognitive labor can lead to poor intellectual ability and a willingness to pursue a task despite difficulty. "We find that AI assistance improves immediate performance, but it comes at a heavy cognitive cost,” the study declares of its findings.
This study, which was conducted by a talented cohort of scientists across the United States and the United Kingdom, has yet to be peer-reviewed. However, it builds on the growing body of scientists and researchers studying the consequences of integrating AI into our daily lives. To conduct the study, the researchers recruited around 350 Americans, who were asked to complete a set of tasks. However, only half were granted access to a chatbot for help.

Everyone else was funneled into an AI-free control group. Initially, the group with access to AI was performing significantly better, as you'd expect. But the tables turned when, suddenly, the access was cut off during the test. "After just [about] 10 minutes of AI-assisted problem-solving, people who lost access to the AI performed worse and gave up more frequently than those who never used it," it further noted. For a follow-up experiment, researchers recruited another larger group of about 670 participants and split them into groups for the test.
The results were pretty much identical to the first one. These same outcomes persisted once again in a final experiment, in which about 200 more participants were asked to complete a brief series of reading comprehension questions, showing that such results aren’t simply limited to math problems. "Once the AI is taken away from people, it’s not that people are just giving wrong answers. They’re also not willing to try without AI," said Rachit Dubey, assistant professor at the University of California, Los Angeles.

Overall, Dubey and his team highlighted that blindly integrating chatbots into educational programs and daily lives could prove to be catastrophic in the near future. Even more so, with global adaptation of generative AI tools already up from 15.1% in the first half of 2025 of the world's population to about 16.3% during the second half. This means that today, roughly one in six people are using AI to learn, work, or solve problems, according to Microsoft.
That said, with this number set to only increase even further, if we keep offloading AI at scale for everything and anything, what will this do to our thinking and beliefs? We might have a generation of learners and people who don't know what they are capable of and are unable to work without Artificial Intelligence. So, as the researchers seek to expand their research into long-term experiments, they're trying to not only learn what people can do with AI, but also what they can do without it.
Cambridge professor explains why you should always say 'please' and 'thank you' to your AI chatbot
AI was asked what dreams actually look like and the results have shocked everyone