AI Security Wiki Update: The Regional AI Regulatory Hub
The AI Security Wiki's biggest update yet: regulatory coverage for 12 regions, 51 templates, 44 control objectives and 629 test cases.
https://nachi065.github.io/AI-Security-Wiki/
AI Security Wiki pages got their biggest update yet today. Regional AI Regulatory Hub.
- 12 sections covering the GCC, EU, UK, US, India, Singapore, Japan, South Korea, China, Canada, Australia, and Brazil.
- Each one lists the laws, guidance, and standards that bear on AI, maps them to security controls, and comes with a role guide and fill-in templates.
- 51 templates in all, from an EU AI Act classification form to a DPDP processing record for India.
- 44 control objectives. Four are new: fairness and bias testing, fundamental rights impact assessment, regulator engagement, and AI literacy. Every control cites OWASP, MITRE ATLAS, NIST AI RMF, and ISO/IEC 42001 identifiers and links to the test cases that exercise it.
- 629 test cases for evaluating AI security products in a proof of concept, each with a procedure, expected results, and pass criteria.
- Sector overlays for financial services, insurance, health and government.
- A Foundations paper on the difference between AI security and security of AI.
A redesigned site with search, a cleaner home page, and country flags in the navigation, so it is easier to find your region.
One thing to be upfront about: the regional legal content is compiled from multiple secondary sources. Every claim is tagged with how it was sourced, and every page says to check the primary text before relying on it. It is a starting point for your own verification, and it is not legal advice.