New research published Tuesday in RSNA’s Radiology explores factors that make radiologists less likely to be fooled by incorrect advice from large language models. South Korean researchers conducted a retrospective study involving 10 readers interpreting chest imaging from 100 patients. They found that successful radiologist-LLM collaboration was associated with confidence in the AI model and reader expertise in chest imaging. The study used curated data from the Korean Society of Thoracic Radiology Weekly Case platform, spanning 2018 to 2020, including X-rays, CTs, MRIs and PET scans. Readers handled one session without AI and a second randomized to use one of two LLM models: a high accuracy (76%) model using OpenAI’s GPT-5 or a low accuracy one (27%) employing GPT-4o. Model confidence (odds ratio of 3.82) and reader expertise (OR, 2.06) were independently associated with adequate radiologist-LLM interaction. The confidence effect was weaker among expert readers (OR, 0.79). Higher rationale quality reduced readers’ rejection of correct suggestions (OR, 0.79) but increased acceptance of incorrect suggestions (OR, 1.71). Higher reader expertise (OR, 0.54) and reader confidence (OR, 0.80) were protective, reducing acceptance of incorrect suggestions. Corresponding author Taehee Lee, MD, MSc, noted that while high model confidence was associated with correct decisions, expertise may safeguard against