What question did this study set out to answer?

This research investigates the effectiveness of generative AI models in determining medical malpractice in fictional clinical cases.

Could ChatGPT and co. replace forensic experts? A comparative study on medical liability expertise

View Full Paper

Ask AI

Bookmark

Authors

Comparative study assesses generative AI performance in medical liability assessments, suggesting limitations in forensic expertise.

This research investigates the effectiveness of generative AI models in determining medical malpractice in fictional clinical cases.
Utilized three generative AI models (ChatGPT-4 Turbo, Gemini, Mistral AI)
Presented nine fictional clinical cases to the AI models
Compared AI conclusions to those of three expert panelists
Conducted five queries for each case, totaling 135 requests
AI models aligned with expert conclusions in 86 out of 135 instances
Discrepancies included false negatives and false positives
All models showed internal contradictions at least once
Current AI models are inadequate for rigorous forensic assessments

Bérar et al. (2026) studied this question.

View Full Paper

Ask AI

Bookmark