AI chatbots reading X-rays can be dangerously confident even when they're wrong

AI models tested on the RadLE 2.0 benchmark frequently deliver incorrect X-ray diagnoses with high confidence, underscoring that human radiologists remain superior. The findings highlight the need for AI systems to recognize when to defer to human expertise before autonomous diagnosis is safe.

The RadLE 2.0 benchmark tests whether AI models in radiology can tell when they should leave a diagnosis to a human. Many models deliver wrong findings with full confidence, and human radiologists are still well ahead. Before AI can diagnose on its own, it needs to learn when it's better to say nothing. The article AI chatbots reading X-rays can be dangerously confident even when they're wrong https://the-decoder.com/ai-chatbots-reading-x-rays-can-be-dangerously-confident-even-when-theyre-wrong/ appeared first on The Decoder https://the-decoder.com .