What are the best AI security frameworks available?

The best AI security frameworks include: OWASP LLM Top 10 for LLM application risks, OWASP Agentic Top 10 for autonomous agent security, NIST AI Risk Management Framework (RMF 1.1) for comprehensive AI risk management, and MITRE ATLAS for adversarial threat knowledge in machine learning systems.

What resources are available for learning LLM security?

Key LLM security learning resources include: the OWASP GenAI project for guides and tools, the OWASP AI Exchange for assessment methodologies, NIST publications on adversarial ML, the AI Hacking Portal for practical pentesting methodology, and curated tool collections for prompt injection scanning and MCP security testing.

Which regulations apply to AI security?

Major AI security regulations include: the EU AI Act requiring risk classification and compliance for high-risk AI systems, NIST AI RMF providing voluntary risk management guidance, ISO/IEC 42001 for AI management systems, and emerging state-level regulations in the US. Organizations should also follow OWASP AI security best practices.

What tools are essential for AI security testing?

Essential AI security testing tools include: Garak and PyRIT for LLM vulnerability scanning, MCP Inspector for MCP server testing, ModelScan for detecting malicious models, PromptInject for injection testing, and the OWASP AI Exchange toolkit for comprehensive security assessment of AI systems.

AI Security Resources

Curated collection of frameworks, guides, tools, and research for AI security testing

OWASP LLM Top 10 2025

Framework

The definitive list of critical LLM application security risks

Visit

OWASP Top 10 for Agentic Applications 2026

Framework

Security risks specific to autonomous AI agents

Visit

NIST AI Risk Management Framework

Framework

US government AI risk management guidelines

Visit

EU AI Act Official Text

Regulation

Comprehensive EU AI regulation

Visit

OWASP AI Security Project

Project

Main OWASP GenAI security resources

Visit

Secure MCP Development Guide

Guide

Securing Model Context Protocol servers

Visit

MITRE ATLAS

Framework

Adversarial threat framework for ML systems

Visit

NIST AI 100-2

Publication

Adversarial machine learning taxonomy

Visit

OWASP AI-Exchange

Tool

AI security assessment tool

Visit

AIBOM Generator

Tool

AI Bill of Materials generation tool

Visit

State of Agentic AI Security and Governance 2.01

Report

Comprehensive landscape report on securing autonomous AI systems, frameworks, and global regulations

Visit

AIUC-1 Crosswalks for Agentic Applications

Framework

Bidirectional mapping between AIUC-1 requirements and the OWASP Agentic Top 10

Visit

AI Security Solutions Landscape for Agentic AI Q2 2026

Report

Maps the agentic AI lifecycle security tools across all deployment stages

Visit

AI Security Solutions Landscape for Agentic Red Teaming Q2 2026

Report

Focused landscape of red teaming tools and frameworks for agentic AI systems

Visit

Research Papers

Training Materials

Contributing

These resources are maintained by the security community. To contribute:

Join the OWASP GenAI Project
Submit issues and pull requests to relevant repositories
Share your research and findings

AI Hacking Team

The AI Hacking team researches and documents AI/LLM security vulnerabilities, red teaming techniques, and defensive strategies. Our guides are based on real-world pentesting experience and continuous monitoring of the AI security landscape.

GitHub LinkedIn