Hey Product Hunt — I am Koukyosyumei, the developer of AIJack, a security and privacy risk assessment tool for machine learning models. Recent studies have shown that machine learning / deep learning models might have severe security/privacy threats such as Model Inversion, Membership Inference, Back Door, Poisoning, and Free Rider. On the other hand, researchers have developed many defense techniques such as Differential Privacy, Homomorphic Encryption, K-Anonymity, and Federated Learning. AIJack allows you to experiment with various attack and defense algorithms combinations. We support more than 30 state-of-the-art methods. AIJack is one of the easy-to-use OSS tools to simulate security and privacy vulnerabilities of machine learning with some unique features: - All-around abilities for both attack & defense - PyTorch-friendly design - Compatible with scikit-learn - Fast Implementation with C++ backend - MPI-Backend for Federated Learning - Extensible modular APIs

AIJack

Security and Privacy Risk Simulator for Machine Learning

Security and Privacy Risk Simulator for Machine Learning