OpenAI's Whisper model is reportedly 'hallucinating' in high-risk situations

This was originally published on post
The tool has widespread adoption in US hospitals, but researchers are warning of potential harms