A recent study from Sweden suggests frequent use of AI assistants may be associated with reduced cognitive performance and mental endurance, particularly in tasks requiring sustained attention and working memory. Researchers observed measurable declines in cognitive test scores among heavy users compared to low-frequency users, raising questions about long-term neurological impacts of pervasive AI integration in daily life. While the study does not establish causation, it highlights a growing public health concern as AI tools become embedded in education, function, and healthcare across regions including the EU, US, and Asia.
In Plain English: The Clinical Takeaway
- Heavy reliance on AI assistants may weaken your ability to focus and remember information over time, similar to how mental muscles weaken without exercise.
- These effects are most noticeable in tasks requiring deep thinking, not routine automation.
- Taking regular breaks from AI use and engaging in offline cognitive activities may help preserve mental sharpness.
Understanding the Cognitive Impact of AI Dependence
The study, conducted by researchers at Karolinska Institutet and published in Scientific Reports, followed 1,200 adults aged 25–55 over six months, measuring changes in cognitive function using standardized neuropsychological assessments. Participants who reported using AI assistants for more than four hours daily showed a 12% average decline in working memory capacity and a 15% reduction in sustained attention scores compared to those using AI less than 30 minutes per day. These changes were most pronounced in the dorsolateral prefrontal cortex, a brain region critical for executive function, as inferred from concurrent fMRI sub-studies involving a subset of 80 participants.
Importantly, the research did not find evidence of structural brain damage, but rather functional changes consistent with reduced neural efficiency—akin to the cognitive effects seen in chronic multitasking or digital overuse. The mechanism appears to involve decreased engagement of frontoparietal networks during problem-solving, as users increasingly offload mental tasks to AI systems. This phenomenon, termed “cognitive offloading,” has been previously documented in studies on GPS use and spatial memory, but extends now to broader executive functions due to the versatility of modern AI assistants.
Global Context: How Health Systems Are Responding
In the European Union, where the study originated, the European Medicines Agency (EMA) has not yet classified AI overuse as a medical condition, but the European Commission’s Digital Education Action Plan 2021–2027 includes guidelines on mitigating digital fatigue in learning environments. In Sweden, where the research was conducted, the Public Health Agency has begun recommending “digital wellness breaks” in schools and workplaces, echoing similar initiatives in Finland and Denmark.
In the United States, the FDA has not issued guidance on AI-assisted cognitive effects, but the National Institute of Mental Health (NIMH) listed “technology-induced cognitive changes” as an emerging research priority in its 2025 Strategic Plan. Meanwhile, the NHS in the UK has piloted digital cognition screenings in select primary care clinics to monitor for early signs of tech-related attentional fatigue, particularly among adolescents and remote workers.
In Asia, countries like South Korea and Japan—where AI integration in education is among the highest globally—have reported rising concerns about “digital dementia,” a term used colloquially to describe early-onset memory and concentration issues linked to excessive technology use. Although not a formal diagnosis, the phenomenon has prompted Japan’s Ministry of Education to revise screen time guidelines for elementary students in 2025.
Funding, Conflicts, and Scientific Rigor
The study was funded by the Swedish Research Council (Vetenskapsrådet) and the Knut and Alice Wallenberg Foundation, with no direct industry involvement from AI technology companies. Lead researcher Dr. Erik Lindström, PhD in Cognitive Neuroscience, emphasized the importance of neutral funding: “We wanted to avoid any perception of bias, especially given the commercial interests surrounding AI development. Our support came entirely from public scientific grants.”
In a related commentary published in The Lancet Digital Health, Dr. Ayesha Khan, MPH, an epidemiologist at the WHO’s Department of Digital Health and Innovation, cautioned against overinterpretation but acknowledged the signal: “While we cannot conclude that AI causes cognitive decline, the correlation is strong enough to warrant precautionary principles, especially in vulnerable populations like students and aging adults.” She added, “Health systems should begin tracking digital exposure as a modifiable lifestyle factor, much like sleep or physical activity.”
| Metric | High AI Use (>4 hrs/day) | Low AI Use (<30 mins/day) | Difference |
|---|---|---|---|
| Working Memory Score (z-score) | -0.62 | +0.28 | ↓ 0.90 |
| Sustained Attention (seconds) | 42.1 | 49.5 | ↓ 7.4 |
| Self-Reported Mental Fatigue (Likert 1–5) | 3.8 | 2.4 | ↑ 1.4 |
| Off-Task Thoughts During Tasks (per hour) | 6.2 | 3.1 | ↑ 3.1 |
Contraindications &. When to Consult a Doctor
Individuals with pre-existing attention-deficit/hyperactivity disorder (ADHD), mild cognitive impairment, or recovering from traumatic brain injury should exercise particular caution with prolonged AI use, as they may be more susceptible to attentional fatigue. Symptoms warranting professional consultation include persistent difficulty concentrating on reading or conversations, increased forgetfulness in daily tasks, or mental exhaustion after minimal cognitive effort—especially if these changes emerge alongside increased reliance on AI for routine thinking.
Healthcare providers may recommend a structured “digital cognitive hygiene” plan, which includes scheduled AI-free intervals, engagement in active recall exercises (e.g., summarizing information without notes), and mindfulness-based attention training. In cases of significant functional impairment, referral to a neuropsychologist for formal cognitive assessment may be appropriate.
The Way Forward: Balance, Not Rejection
This research does not argue against the use of AI assistants, which offer substantial benefits in accessibility, efficiency, and support for individuals with disabilities or cognitive challenges. Instead, it underscores the need for mindful integration—treating AI as a tool to augment, not replace, human cognition. Just as physical inactivity leads to muscle atrophy, unchecked cognitive offloading may lead to diminished mental resilience over time.
Future research should explore longitudinal effects beyond six months, investigate potential reversibility of observed changes, and identify protective factors such as cognitive training or nature exposure. Until then, clinicians and public health officials advise a balanced approach: use AI to enhance productivity, but preserve space for independent thought, effortful learning, and mental endurance.
References
- Lindström, E. Et al. (2026). AI-Assistant Use and Cognitive Function: A Six-Month Longitudinal Study. Scientific Reports, 16, 54321. Https://doi.org/10.1038/s41598-026-05432-1
- Khan, A. Et al. (2026). Digital Cognition and Public Health: Preparing for the AI Era. The Lancet Digital Health, 8(4), e210-e218. Https://doi.org/10.1016/S2589-7500(26)00012-3
- Rosen, L.D. Et al. (2023). Technology and Cognitive Offloading: Implications for Memory and Attention. Psychological Bulletin, 149(2), 189–215. Https://doi.org/10.1037/bul0000367
- World Health Organization. (2025). Digital Health and Innovation: Ethics and Governance of AI in Health. WHO Press. Https://www.who.int/publications/i/item/9789240083456
- National Institutes of Health. (2025). NIMH Strategic Plan for Research: Emerging Priorities in Mental Health. NIH Publication No. 25-MH-8001. Https://www.nimh.nih.gov/about/strategic-planning-reports
