Don't train medical AI on patients' data without their knowledge
Briefly

Uncontrolled deployment of artificial intelligence in the medical field poses risks to electronic health records by potentially contaminating data. Such contamination could lead to distortions in the records, which might cause long-term damage to models designed to interpret that data. Moreover, there is a lack of transparency for individuals about how their health records are being used to train and refine AI tools, contributing to a misunderstanding of the ramifications of AI integration into healthcare.
Uncontrolled deployment of artificial intelligence in medicine could lead to contamination of electronic health records, resulting in data distortions and long-term deterioration in data interpretation models.
A significant concern is that individuals often lack transparency regarding how their health records are utilized to develop or improve AI tools.
Read at Nature
[
|
]