Explainable artificial intelligence (XAI) provides more information to help users to understand model decisions, yet this extra knowledge exposes additional risks for privacy attacks . We study this risk for image-based modelinversion attacks and identified several attack architectures with increasingperformance to reconstruct private image data from model explanations . These threats highlight the urgent and significant privacy risks of explanations and calls attention for new privacy preservation techniques that balance the dual requirement forAI explainability and privacy . The authors conclude that these threats highlightthe urgent . and significant . privacy risks highlight the . urgent and . significant privacy . risks of explaining AI models and calling attention to new privacy . preservation techniques to balance the . dual-requirement for explainable AI explainability . and privacy. For more information, please visit the authors’s website: http://www.jenn.com/jennennennarjornarjorjornor

Author(s) : Xuejun Zhao, Wencan Zhang, Xiaokui Xiao, Brian Y. Lim

Links : PDF - Abstract

Code :
Coursera

Keywords : privacy - risks - explanations - model - urgent -

Leave a Reply

Your email address will not be published. Required fields are marked *