Artificial intelligence tools such as ChatGPT, Claude, and Gemini have become part of everyday life. They draft emails, summarise documents, answer questions, write code, and even help students complete assignments in minutes. But as AI becomes more capable, neuroscientists are raising an uncomfortable question: Could relying too heavily on AI weaken the human brain over time?

While there is no evidence that AI directly causes dementia, a growing body of research suggests that excessive dependence on AI for thinking, remembering, and problem-solving could reduce the mental exercise the brain needs to stay healthy. Experts warn that this phenomenon—known as cognitive offloading—may have long-term consequences if it becomes a habitual replacement for independent thinking.

The concern isn’t that AI exists. It’s how people choose to use it.

TL;DR

• Experts are warning that excessive reliance on AI may reduce the brain’s natural cognitive activity.
• Scientists describe this behaviour as cognitive offloading, where mental tasks are delegated to technology.
• Research suggests regular mental effort strengthens neuroplasticity—the brain’s ability to form new neural connections.
• One recent study found significantly lower brain activity among students who used ChatGPT to write essays compared to those relying on books or lectures.
• Researchers stress that AI should supplement human thinking, not replace it.

What Is Cognitive Offloading?

Cognitive offloading refers to shifting mental work from the brain to an external tool.

Humans have always done this to some degree. Writing grocery lists instead of memorising them or using calculators for complex arithmetic are common examples. AI, however, expands cognitive offloading far beyond memory or calculations.

Today, people increasingly ask AI to do the following:

• Draft emails
• Write reports
• Summarize articles
• Solve technical problems
• Brainstorm ideas
• Explain concepts
• Generate code
• Make decisions

The more AI performs these tasks, the less the brain practises the skills required to perform them independently.

That shift is what concerns neuroscientists.

Why Does Mental Exercise Matter for Brain Health?

The brain functions much like a muscle—it becomes stronger through consistent use.

Neuroplasticity Depends on Active Thinking

Every time someone solves a problem, learns a language, remembers information, or analyses a complex issue, the brain strengthens neural pathways through a process called neuroplasticity.

These repeated mental challenges help the brain:

• Form new neural connections
• Strengthen existing networks
• Improve memory retention
• Enhance reasoning abilities
• Maintain cognitive flexibility as people age

Conversely, if those pathways are used less frequently, they can gradually weaken.

Researchers say this principle explains why activities like reading, learning new skills, solving puzzles, and engaging in meaningful discussions are consistently linked with healthier cognitive aging.

What Researchers Found About AI and Brain Activity

Several recent studies have added weight to concerns surrounding AI dependence.

Italian Neurologist Warns of Passive AI Dependence

In January, Simone Rossi, Professor of Neurology at the University of Siena in Italy, argued that excessive dependence on AI—even for simple activities such as writing emails—could reduce activity-dependent neuroplasticity.

According to Rossi, passive reliance on AI may gradually weaken the brain’s natural ability to strengthen and reorganize neural networks through repeated cognitive effort.

His warning focuses less on AI itself and more on uncritical dependence.

In other words, asking AI for assistance occasionally is very different from outsourcing nearly every thinking task.

MIT Study Found Lower Brain Activity Among ChatGPT Users

One of the most talked-about findings comes from research led by Nataliya Kosmyna, a research scientist at the Massachusetts Institute of Technology.

Kosmyna divided students into three groups and asked each to write an essay.

The groups included:

• Students using ChatGPT
• Students using Google Search
• Students relying only on lectures and books

Researchers then examined brain activity.

The findings were striking.

Students Using Books Showed the Highest Brain Engagement

Students who relied on traditional learning materials demonstrated the highest levels of brain activity across multiple regions associated with learning, memory, and reasoning.

Those using Google showed increased activity in regions linked to visual processing and information retrieval.

Students using ChatGPT exhibited approximately 55% lower brain activity than students who worked without digital assistance.

Perhaps more concerning, many of these students reportedly struggled to remember what they had written only days after submitting their essays.

While one study cannot establish long-term cognitive decline, it raises important questions about how AI-generated work affects memory formation.

Why Learning Something Yourself Creates Better Memory

Memory isn’t created simply by reading information.

It develops through effort.

Educational psychology has long shown that actively working through problems produces stronger long-term memory than simply receiving answers.

This phenomenon is sometimes called desirable difficulty—the idea that learning becomes more durable when the brain has to struggle slightly before arriving at a solution.

AI can remove much of that struggle.

If students immediately receive polished answers instead of analysing concepts themselves, they may remember less because the brain never completed the deeper processing required for lasting learning.

Classroom Experiences Are Raising Similar Concerns

Some educators say they’re already seeing signs of this shift.

A professor at Brown University recently described an unusual pattern among students.

When midterm exams were completed remotely with access to online resources, average scores remained relatively high.

However, when students later took an in-person final examination without those tools, many struggled significantly.

Several students reportedly scored zero, while only a small number performed similarly to their earlier results.

Although these observations do not prove AI caused poorer performance, they suggest that dependency on digital assistance may create gaps between assisted performance and actual understanding.

Could Heavy AI Use Increase Dementia Risk?

This is where the discussion requires careful nuance.

Experts Say the Risk Is Theoretical, Not Proven

Some neuroscientists believe widespread AI dependence could eventually contribute to higher rates of cognitive decline if it consistently reduces mental stimulation across large populations.

Barbara Sahakian, Professor of Clinical Neuropsychology at the University of Cambridge, has argued that maintaining active neural circuits is essential for preserving cognitive function throughout life.

According to Sahakian, keeping the brain engaged helps reduce dementia risk because neural networks remain active and adaptable.

However, it’s important to distinguish between risk factors and direct causes.

Current research does not demonstrate that ChatGPT, Claude, Gemini, or any AI tool causes dementia.

Instead, experts suggest that replacing regular cognitive effort with AI over many years could become one lifestyle factor among many that influence long-term brain health.

Other well-established dementia risk factors include:

• Physical inactivity
• Poor cardiovascular health
• Smoking
• Diabetes
• Chronic sleep deprivation
• Social isolation
• Limited lifelong learning

AI dependence has not yet reached the same level of scientific evidence.

Does This Mean People Should Stop Using AI?

Not at all.

Most researchers emphasize that AI is best viewed as a tool—not a substitute for thinking.

Healthy AI use may include:

• Using AI to brainstorm rather than write everything.
• Asking AI to explain concepts instead of simply providing answers.
• Solving problems independently before checking AI’s solution.
• Editing AI-generated drafts instead of copying them unchanged.
• Continuing to read books, learn new skills, and practise critical thinking.

The goal is collaboration, not replacement.

Just as calculators did not eliminate mathematics education, AI does not have to eliminate independent reasoning.

The Bigger Question Is About Human Habits

The debate over AI and brain health reflects a broader challenge facing modern society.

Technology has steadily reduced the need to memorise phone numbers, navigate using maps, or perform mental arithmetic. AI now extends that convenience into writing, reasoning, creativity, and decision-making.

Convenience is valuable.

But convenience also comes with trade-offs.

If AI becomes the default answer to every cognitive challenge, people may gradually lose opportunities to exercise the very mental abilities that support lifelong learning and resilience.

Researchers say the healthiest future is likely one where AI amplifies human intelligence rather than replacing it.

The Bottom Line

Artificial intelligence is transforming how people work, study, and communicate. Its productivity benefits are undeniable.

At the same time, emerging research suggests that overusing AI for tasks requiring memory, analysis, and problem-solving could reduce the mental effort necessary to keep the brain adaptable.

Scientists are not arguing that ChatGPT or other AI assistants cause dementia. Instead, they are urging caution against excessive cognitive offloading that leaves the brain underused.

As research continues, the message remains simple: use AI to enhance your thinking—not to stop thinking altogether.