Securing AI—and using AI for security
These are related but different markets. One protects models, AI applications, and agents. The other adds AI to established security work such as detection, investigation, and remediation. The registry marks both roles explicitly so a product name or marketing claim never has to do that job.
AI security lifecycle
Three categories, from inventory to runtime
AI Security Posture & Governance
Inventories AI models, applications, agents, data flows, and services; maps their ownership and risk; and helps teams govern AI use against policy and control frameworks.
Learn and browseAI Security Testing & Model Assurance
Tests AI models and applications for prompt injection, jailbreaks, data leakage, unsafe behavior, and model supply-chain risks before and after deployment.
Learn and browseAI Application & Agent Security
Protects AI applications and agents at runtime by inspecting prompts, responses, retrieved content, tool calls, and data flows, then enforcing policy before unsafe actions complete.
Learn and browseFeatured registry entries
Commercial platforms and open projects
Put it into practice
An AI overlay, not a replacement stack
AI-specific controls sit beside identity, application security, data protection, monitoring, governance, and recovery. The curated AI product-team stack shows how those pieces work together, while the AI-enabled rationalization profile adds the NIST AI RMF without penalizing organizations that do not materially use AI.