Doctors Lose Skill After Relying on AI in Colonoscopies

Study Finds Doctors Lose Skill After Relying on AI in Colonoscopies

Written by Aaron S.,

Editor-In-Chief

Last Updated: August 19, 2025

Key Takeaways

Polish study found gastroenterologists’ polyp detection dropped from 28.4% to 22.4% after relying on AI during colonoscopies;
AI software helped spot abnormalities in real time, but doctors’ performance fell once the system was switched off;
Researchers suggest doctors may wait for AI prompts instead of scanning footage themselves.

Ace quick missions & earn crypto rewards while gaining real-world Web3 skills. Participate Now! 🔥

BITDEGREE IN YOUR SOCIAL FEED

A study in Poland has shown that gastroenterologists became less effective at spotting abnormalities after they grew used to working with artificial intelligence (AI) during colonoscopies.

According to an August 19 report by National Public Radio (NPR), the research was carried out at four clinics where doctors tested a system that reviewed live video and marked suspicious areas in real time.

When the software highlighted a region, doctors could check it immediately. The system was successful while in use, but the study also revealed an unintended effect.

What is SushiSwap? DEX & Sushi Token Animated Explainer

Did you know?

Want to get smarter & wealthier with crypto?

Subscribe - We publish new crypto explainer videos every week!

After relying on the tool, doctors' ability to detect possible polyps fell from 28.4% before the trial to 22.4% once the AI was switched off. This means that detection rates dropped by about one-fifth when the automated help was removed.

The findings were published in The Lancet Gastroenterology and Hepatology.

Lead researcher Marcin Romańczyk, a gastroenterologist at H-T Medical Center in Tychy, said the results came as a surprise. He noted that many specialists were trained through textbooks and mentorship, but not in using advanced technology like AI, which is spreading through healthcare.

Romańczyk suggested that one possible reason for the drop is that clinicians may unconsciously wait for the system to mark a suspicious area, instead of carefully scanning the footage themselves.

Anthropic recently added a new option to certain Claude models that lets them close a chat in very limited cases. What does the feature include? Read the full story.