Jiankai Jin

New Robust Privacy method protects AI inference from data leakage

Researchers have introduced Robust Privacy (RP), a new method to protect sensitive information during AI model inference. RP leverages certified robustness to ensure model predictions remain invariant within a certain r…