Artificial intelligence is increasingly taking on a wide range of tasks, from writing emails and programming to translation and trip planning, raising questions about the potential long-term decline of human cognitive abilities.

The emergence of generative AI chatbots like ChatGPT, capable of creating various types of content in response to simple requests in everyday language, has transformed usage patterns in schools, workplaces, and daily life.

Recent scientific research involving limited numbers has shown that relying on AI for certain tasks may have negative repercussions on memory, decision-making ability, and critical thinking.

The words 'artificial intelligence,' a keyboard, and robotic hands in this illustration (Reuters)

In April, a US-UK study still under review found that using these tools to solve math problems or complete reading comprehension exercises improved participants' short-term performance but negatively affected their long-term performance and perseverance when AI support was withdrawn.

The authors of the study, conducted on 1,222 people, wrote: 'These findings raise significant concern, as perseverance is a key element in skill acquisition and one of the best indicators of long-term learning.'

In an interview with Agence France-Presse, lead author Grace Liu explained that AI, praised for its speed in calculations, conditions people to expect immediate answers, thereby 'depriving them of learning opportunities.'

She added: 'The worrying thing is that AI is not limited to performing a specific task; it can be employed in various intellectual activities that rely on analysis and deduction,' unlike a calculator, which helps solve equations but leaves the thinking process to the user.

'Effort saving'

A 2025 study by the Massachusetts Institute of Technology showed that students who use generative AI to write essays are less capable of critical thinking.

Other studies support this finding, highlighting a phenomenon known as 'cognitive offloading' or even 'mental disengagement.'

Johan Chevalier, a researcher at the Laboratory of Social and Cognitive Psychology at the French National Centre for Scientific Research, told Agence France-Presse: 'Humans have a strong tendency to save effort.'

He added: 'In our daily lives, we often use strategies that get us to the goal more quickly, without necessarily delving into the information to be processed, which requires significant cognitive effort,' noting that generative AI reinforces this tendency.

He continued: 'If there are activities that a person never engages in, their brain, which operates on the principle of energy conservation, will not bother maintaining unused neural connections.'

AI applications (AFP)

Stimulating thinking

To mitigate these effects and reduce criticism, companies specializing in AI models have developed educational tools based on the Socratic method, targeting students in particular.

Chatbots do not generate automatic answers but provide hints and ask questions to stimulate thinking, such as the 'Study' feature in ChatGPT or the 'Guided Learning' feature in Gemini.

Microsoft stated that it has added alerts about potential errors, reminders to verify information, and various measures to encourage users to actively and critically engage with the answers generated by the tool.

The phrase 'artificial intelligence' (AFP)

Microsoft noted that 'the risk of over-reliance on cognitive offloading exists, especially when AI is used to automate tasks that help develop skills,' stressing the importance of training users to handle these tools.

Large-scale, long-term studies are still needed to determine the actual impact of this new technology on human brains, researchers affirmed.

Chevalier said: 'It is up to us to use AI intelligently,' even if that requires effort, adding: 'We will adapt to this technological revolution just as we have adapted to previous ones.'