A curated list of MLSecOps tools, articles and other resources on security applied to Machine Learning and MLOps systems.
Simple pytorch implementation of FGSM and I-FGSM
pretrained BERT model for cyber security text, learned CyberSecurity Knowledge
CTF challenges designed and implemented in machine learning applications
Code for our USENIX Security 2021 paper -- CADE: Detecting and Explaining Concept Drift Samples for Security Applications
Reading list for adversarial perspective and robustness in deep reinforcement learning.
Adversarial Machine Learning (AML) Capture the Flag (CTF)
Train AI (Keras + Tensorflow) to defend apps with Django REST Framework + Celery + Swagger + JWT - deploys to Kubernetes and OpenShift Container Platform
AI SBOM: AI Software Bill of Materials - The Supply Chain for Artificial Intelligence
Do you want to learn AI Security but don't know where to start ? Take a look at this map.
Hands-on lessons for attacking and defending AI systems, starting with the OWASP Top 10 for LLM Applications.
Honest-but-Curious Nets: Sensitive Attributes of Private Inputs Can Be Secretly Coded into the Classifiers' Outputs (ACM CCS'21)
Datasets for training deep neural networks to defend software applications
Summary of the presentation on Real and Stealthy Attacks on State-of-the-Art Face Recognition Systems at the Seminar: Machine Learning in Cyber-security at FU Berlin
Systematic Security Evaluation Framework for AI Coding Assistants - Detection of prompt injection vulnerabilities
A high-security Linux OS for ethical hacking & penetration testing. Built on Kali & Parrot OS. Features AI intrusion detection, stealth VPN, custom kernel & more.
Educational research demonstrating weight manipulation attacks in SafeTensors models. Proves format validation alone is insufficient for AI model security.
Awesome-DL-Security-and-Privacy-Papers
Veritensor is the Zero-Trust security platform for the AI Supply Chain. We replace naive scanning with deep AST analysis and cryptographic signing. From CI/CD to Kubernetes runtime, Aegis ensures only verified, safe, and compliant models ever reach production. Stop guessing, start proving.
Adversarial perturbation intensity strategy achieving chosen intra-technique transferability level for logistic regression
Add a description, image, and links to the machine-learning-security topic page so that developers can more easily learn about it.
To associate your repository with the machine-learning-security topic, visit your repo's landing page and select "manage topics."