Skip to content
Submit incident
AII-1000-32949
ConfirmedA court, a regulator or the organization itself has stated the core facts, or the record cites a court or regulator source.Supporting sources are pending verification.

AI system causes skill atrophy in medical specialists, Polish study finds

A study of Polish endoscopy centers found that specialists who relied on AI-assisted lesion detection experienced measurable skill decline when the system was removed, with adenoma detection rates dropping 6% despite extensive prior experience. The phenomenon, termed 'deskilling,' suggests that AI assistance can erode professional cognitive habits and practical skills even as it improves immediate performance.

15 July 2026 Β· Safety and wellbeing
15 July 2026
AII-1000-32949
Quick facts
Category
Safety and wellbeing
Country
Poland
Date
15 July 2026
Drafted by
Raidu Incident Detection

What happened

A team of researchers at four endoscopy centers in Poland studied the effect of AI-assisted lesion detection tools on specialist performance. After endoscopists worked with the AI system for a period, researchers evaluated how the same specialists performed when operating without the assistance.

The study, published in The Lancet Gastroenterology & Hepatology, found a 6% absolute decline in adenoma detection rates among nineteen endoscopists when the AI system was removed, despite their extensive prior experience with thousands of procedures.

Reported impact

The specialists experienced atrophy of professional skills, specifically losing the habit of careful visual inspection during procedures, even though their eyesight remained intact.

Index analysis

The phenomenon, termed 'deskilling,' represents a long-term consequence of overreliance on AI assistance where workers lose cognitive and practical skills due to reduced need to perform critical thinking tasks.

The findings suggest that AI assistance, while improving immediate performance metrics, can erode the muscle memory and attention patterns that experts rely on when the system is unavailable.

Governance controls

The study indicates that organizations deploying AI assistance in high-stakes medical fields may not be adequately managing the risk of skill degradation among their staff, despite investments in training and reskilling programs.

Classification

Category
Safety and wellbeing
Industry
Healthcare, drugs and biotechnology
Country
Poland
Harm status
Actual harm
Intent
Unintentional
Timing
Post-deployment
Harm types
Psychological
Harmed
Workers
Failure modes
Overreliance
AI systems
AI lesion detection system

Sources and evidence

Sources not yet published

This record was researched and written by the Index from original reporting. Its sources are still being verified by an editor, and they are listed here once that check is complete. Until then the record carries no reviewer.

How sources are verifiedSuggest a source for this incident