From 41082ce1ef0da8338905ab823096ecf162a08214 Mon Sep 17 00:00:00 2001 From: Matteo Meucci Date: Sun, 23 Nov 2025 17:39:00 +0100 Subject: [PATCH] Update AITG-MOD-05_Testing_for_Inversion_Attacks.md --- .../AITG-MOD-05_Testing_for_Inversion_Attacks.md | 16 ++++------------ 1 file changed, 4 insertions(+), 12 deletions(-) diff --git a/Document/content/tests/AITG-MOD-05_Testing_for_Inversion_Attacks.md b/Document/content/tests/AITG-MOD-05_Testing_for_Inversion_Attacks.md index 4306020..bbdb3a3 100644 --- a/Document/content/tests/AITG-MOD-05_Testing_for_Inversion_Attacks.md +++ b/Document/content/tests/AITG-MOD-05_Testing_for_Inversion_Attacks.md @@ -30,18 +30,10 @@ This test identifies vulnerabilities associated with model inversion attacks, wh - **Regular Privacy Audits**: Regularly perform model inversion attacks against your own models as part of a security audit to proactively identify and mitigate vulnerabilities. ### Suggested Tools for this Specific Test -- **Adversarial Robustness Toolbox (ART)** - - Includes implementations of various model inversion attacks, allowing you to test your model's susceptibility. - - Tool Link: [ART on GitHub](https://github.com/Trusted-AI/adversarial-robustness-toolbox) -- **TensorFlow Privacy** - - A library for training models with Differential Privacy, which is a primary defense against inversion attacks. - - Tool Link: [TensorFlow Privacy on GitHub](https://github.com/tensorflow/privacy) -- **Opacus (for PyTorch)** - - A library from Meta that enables training PyTorch models with differential privacy. - - Tool Link: [Opacus on GitHub](https://github.com/pytorch/opacus) -- **PrivacyRaven** - - A framework from Trail of Bits specifically designed for privacy testing of deep learning models, including model inversion. - - Tool Link: [PrivacyRaven on GitHub](https://github.com/trailofbits/PrivacyRaven) +- **Adversarial Robustness Toolbox (ART)**: Includes implementations of various model inversion attacks, allowing you to test your model's susceptibility - [ART on GitHub](https://github.com/Trusted-AI/adversarial-robustness-toolbox) +- **TensorFlow Privacy**: A library for training models with Differential Privacy, which is a primary defense against inversion attacks - [TensorFlow Privacy on GitHub](https://github.com/tensorflow/privacy) +- **Opacus (for PyTorch)**: A library from Meta that enables training PyTorch models with differential privacy - [Opacus on GitHub](https://github.com/pytorch/opacus) +- **PrivacyRaven**: A framework from Trail of Bits specifically designed for privacy testing of deep learning models, including model inversion - [PrivacyRaven on GitHub](https://github.com/trailofbits/PrivacyRaven) ### References - Fredrikson, Matt, Somesh Jha, and Thomas Ristenpart. "Model Inversion Attacks that Exploit Confidence Information and Basic Countermeasures." ACM CCS 2015. [Link](https://dl.acm.org/doi/10.1145/2810103.2813677)