Does AI make our brains smarter or lazier?

Forty percent of young people use generative AI every day, barely three years after the launch of ChatGPT. Strikingly, 83% are unable to recount a text they have just generated. This raises important questions. Does the principle “use it or lose it” apply here? Are we risking a loss of brain capacity or even less optimal development, when children ask ChatGPT everything?

So far, little research has been conducted on the impact of AI on our brains. Should we fear negative consequences?

Our brains continuously form new connections between nerve cells throughout life, a unique ability known as plasticity. This plasticity depends on activity and must be constantly stimulated. A baby, for instance, learns to speak through repetition and practice, gradually creating lasting networks between language and memory. Thanks to this plasticity, we can keep learning throughout life, recover after brain damage, and even slow down dementia.

Recent neuroscientific research now suggests that regular use of AI for text production may weaken certain brain connections, particularly in the networks responsible for language and memory. When AI is used, information is processed less deeply, and AI-generated texts are harder for us to remember. In the long run, these networks may function less effectively and become less capable of storing new information. In the scientific literature, this is sometimes referred to as the accumulation of “cognitive debt” when we outsource mental effort to AI. There is therefore concern that our brains may become lazy, with important neural networks gradually weakening or even disappearing. This raises worries about a possible acceleration or increased risk of cognitive decline, such as dementia, something we are trying to prevent in our ageing societies.

At the same time, AI can make certain cognitive processes faster and more energy-efficient, which may also have positive effects. For some people, AI opens up new possibilities, such as developing a business model or generating marketing strategies. AI use may also stimulate our brains. Creativity and cognitive engagement often arise when people carefully question and critically evaluate AI generated results. The mental energy saved can then be redirected to other tasks.

This can also be beneficial in the context of neurological conditions. Patients who survive a stroke but lose some independence may benefit from an AI assistant such as the open-source agent OpenClaw or Claude. Thanks to its “persistent memory” function, such an assistant can learn from a user’s behaviour and perform personalised tasks such as sending emails, managing calendars, searching online and using digital services.

The growing interaction between humans and AI is irreversible. The concern is that we may begin to rely on AI passively and uncritically, both in general and in tasks that require cognitive engagement. This could weaken the plasticity of our brains. It may also threaten our collective capacity for thinking and creativity. Ultimately, it remains our responsibility to give meaning to the results generated by AI. Doing so stimulates cognitive engagement and provides the mental energy that fuels brain plasticity.

As AI and large language models become increasingly present in cognitive tasks, we will need to actively focus on co-creation in order to preserve and even strengthen our brain plasticity. This requires a preventive framework. We need what might be called “cognitive hygiene.” AI use in education and everyday tasks can be encouraged, but the meaning of it all must remain central. Only in this way can we safeguard our brain health in the long term, both as individuals and as a society.