A new study in The Lancet Gastroenterology and Hepatology has highlighted the potential downside of integrating artificial intelligence (AI) in routine medical practice, raising questions about whether overreliance on decision-support systems could diminish clinicians’ core skills.
According to The Economic Times, the research examined colonoscopy procedures at four centres in Poland, where AI tools for detecting polyps and abnormal cells were introduced in late 2021. The study found that the adenoma detection rate (ADR) in non-AI-assisted colonoscopies declined from 28 per cent before AI exposure to 22 per cent after AI exposure. This drop, representing a 20 per cent relative and 6 per cent absolute reduction, suggests that clinicians may have been less effective when operating without AI assistance after regular exposure to such systems.
Researchers described this as the first documented evidence of a potential negative impact of AI on healthcare professionals’ ability to perform a critical medical task. “Our results are concerning given the adoption of AI in medicine is rapidly spreading. We urgently need more research into the impact of AI on health professionals’ skills across different medical fields,” said Dr. Marcin Romarnczyk of the Academy of Silesia, who co-authored the study.
Co-author Yuichi Mori of the University of Oslo noted that while earlier trials suggested AI-assisted colonoscopy improved ADR, those studies might not have accounted for potential deskilling effects. Continuous exposure to automated recommendations, the authors argue, may lead clinicians to unconsciously rely more on technology than on their own judgment, reducing focus and motivation during procedures.
The study was supported by the European Commission, the Japan Society for the Promotion of Science, and the Italian Association for Cancer Research. Its findings underscore a key challenge for healthcare systems adopting AI: balancing the efficiency and precision of advanced tools with the need to preserve and enhance human expertise.
Industry stakeholders have responded with calls for a nuanced perspective. Dr. Vidur Mahajan, Founder and CEO of CARPL.AI, suggested the focus should remain on how technology can raise the performance of average clinicians to global standards rather than the fear of skill erosion. Drawing a parallel, he remarked: “Technology is an inevitable part of our lives and we must embrace the advantages of it by enabling its democratisation. Imagine a world without Google Maps – would you trust a driver who does not use it?”
For policymakers, hospital administrators, and MedTech developers, the findings highlight the importance of designing AI integration strategies that safeguard clinical skills. Structured training programs, periodic skill assessments, and hybrid decision-making models may be essential to ensure that technology complements rather than substitutes medical expertise.
