In a startling revelation that threatens to upend the burgeoning love affair with AI assistants, researchers at MIT’s Media Lab warn that heavy reliance on chatbots like ChatGPT may inflict lasting damage on our cognitive faculties.
In a four-month experiment involving 54 volunteers, subjects were split into three cohorts—those using ChatGPT, those using Google Search and a “brain-only” group relying purely on their own knowledge.
Electroencephalography (EEG) scans showed that the ChatGPT users exhibited the weakest neural connectivity and engagement across alpha and beta bands, underperforming at neural, linguistic and behavioural levels by the study’s close.
Rather than deepening critical thought, the LLM cohort gradually abandoned probing questions, resorting instead to copy-pasting AI-generated prose.
By contrast, Google users demonstrated moderate brain activity, while the brain-only group forged the strongest and most distributed neural networks—proof, critics argue, that outsourcing thought to algorithms comes at a steep price.
Lead author Nataliya Kosmyna sounded the alarm for policymakers, especially those eyeing AI integration in early education:
“I am afraid in 6–8 months, there will be some policymaker who decides, ‘let’s do GPT kindergarten.’ I think that would be absolutely detrimental,” she told TIME.
The implications are profound for Nigeria’s tech-savvy youth. With ChatGPT boasting nearly 800 million weekly active users worldwide as of June 2025, including a significant proportion under 25, the risk of an “AI generation” with blunted analytical skills has never been greater.
This is not the first time new technologies have sparked fear over “mental atrophy”. From Socrates’ 4th-century BCE lament that writing would “implant forgetfulness” in souls, to modern studies showing GPS overreliance erodes spatial memory, history is littered with cautionary tales of convenience sapping capability.
Yet the sheer scale of LLM adoption renders today’s warning uniquely urgent.
Educators and policymakers in Nigeria must weigh the seductive efficiencies of AI tools against the peril of cognitive complacency.
As institutions debate whether to ban or embrace ChatGPT, this MIT study serves as a clarion call: without stringent guidelines and balanced usage, we risk crafting a generation fluent in algorithms but atrophied in thought.
