A study summarized in the excerpt describes evidence that clinicians’ performance can decline when AI decision-support tools are unavailable, illustrating a potential “deskilling” risk in medicine. The research followed physicians specializing in endoscopy who were given intermittent access to an AI system that analyzes colonoscopy images in real time and flags adenomas, described as precancerous intestinal lesions. According to the excerpt, when physicians used the tool, their adenoma detection rate dropped significantly on days without AI assistance. The reported adenoma detection rate fell from 28.4% during the three-month period before AI introduction to 22.4% afterward in colonoscopies without AI. The findings were contextualized with broader survey data in which 70% of nurses and 77% of physicians reported worry about losing skills due to over-reliance on AI systems. The excerpt also quotes information scientist Kevin Crowston at Syracuse University, framing the issue as a prompt for “self-reflection” on which skills teams want to retain. For biotech and healthcare technology developers, the practical takeaway is that deployment plans need to consider human factors, training, and fallback performance when AI is removed or fails—especially in regulated clinical settings.