Deepfake X-rays Fool Doctors and AI Models

A study published in the journal Radiology reveals that experts cannot reliably identify deepfake medical images ^[1]. Researchers tested experienced radiologists and cutting-edge large language models against AI-generated X-rays ^[2]. Neither the humans nor the digital models could consistently tell the difference between authentic scans and deepfakes ^[3]. These findings expose serious vulnerabilities in healthcare data security ^[1]. While AI technology offers clinical benefits, it also enables the creation of highly realistic fraudulent records ^[2]. Experts suggest that protecting the integrity of medical data is now a critical challenge for the industry ^[3].