Doctors are worse at detecting cancer symptoms after relying on AI, study finds

The Lancet medical journal has published the results of a study that found a decrease in doctors' effectiveness in detecting colon cancer after using artificial intelligence (AI) tools. The experiment was conducted in four endoscopy centers in Poland for six months: three months before the introduction of AI and three months after.

After the AI integration, colonoscopies were randomly assigned to either with or without AI. It was found that doctors who performed the procedures without AI after the AI integration period had 20% worse outcomes than before the technology was implemented. The study involved 19 experienced doctors, each of whom had performed more than 2,000 colonoscopies.

The study authors note that even highly skilled professionals may lose their professional skills due to overreliance on AI, raising further concerns about the potential consequences for less experienced healthcare professionals.

Although AI demonstrates positive results in medical diagnostics, in particular in the detection of oncological diseases, the study still highlights the risk of reducing critical thinking and professional competence among specialists.

As Gizmodo writes, similar findings were previously made by researchers from the Massachusetts Institute of Technology and Microsoft, who found that users who often rely on AI analyze information less on their own. In this regard, there is a great risk in the future that the overuse of AI tools can dull people's logical analysis and decision-making skills.

According to the American Medical Association, about two-thirds of doctors are already using AI in their practices.