TalEliyahu/Awesome-AI-Security

It's a single, well-organized hub that collects best practices, tools, research, and learning materials — all focused on making AI and machine learning systems safe from attacks and misuse.

What's inside? The repository is split into many categories, so you can jump straight to what you need:

• Best Practices, Frameworks & Controls — the biggest section. It links to official guides and standards from organizations like NIST, OWASP, and the EU, covering topics like risk management, verification, red teaming, threat modeling, and policy templates. There are also self-assessment toolkits and maturity models.
• Tools and Models — practical software and AI models that help with security tasks.
• Datasets — for phishing, deepfakes, jailbreak prompts, and more, useful for testing.
• Learning — courses, certifications, books, videos, podcasts, and training.
• Incident Response — repositories of known vulnerabilities, guides, and playbooks.
• Research — vendor reports, papers, and ongoing working groups.
• Community — links to forums, blogs, and social groups.

Who is it for? Anyone interested in AI security — from developers and security professionals to researchers and policy makers. Because it's a curated list, you don't have to search the web yourself; everything is already organized and vetted by the community.

Organization and maintenance The list is managed by a LinkedIn group called AISecHub and is sponsored by InnovGuard, a cybersecurity advisory firm. The structure uses a clear table of contents, so you can navigate quickly.

In short, it's a one-stop reference for learning about and practicing AI security — a living document that aggregates the most useful resources in the field.