Privacy-preserving learning techniques are fusion techniques that help preserve user privacy. They can use cryptography and statistics tricks. These conflicting needs can be met with a technique called differential privacy, or privacy-preserving AI. The PPAI allows organizations to collect information about their users without compromising an individual's privacy.
The wide application of artificial intelligence techniques in medicine is currently hampered by the limited availability of data sets for training and validation of algorithms, due to the absence of standardized electronic medical records and to strict legal and ethical requirements to protect patient privacy.
