AgentGuardJoin early access

Free Risk Score · No signup

Know where your AI controls need work.

Answer 10 questions. Get a 0–100 readiness score and up to 3 Findings, each with a concrete Fix.

Higher is safer. This is a self-reported governance indicator, not a security test or compliance determination.

Loading your saved answers.

See how scoring works

9 controls total 100 points. The EU context question has no weight.

A complete control earns its full weight. A partial control earns half. A missing or unknown control earns 0.

We add earned points, then round to the nearest whole number. Weights are AgentGuard’s priorities, not a validated standard.

Findings are the 3 largest unearned point gaps. Ties follow questionnaire order. The score reflects reported controls, not measured risk.

  • Do you have a list of the AI tools and agents your team uses?12 points
  • Does each AI tool and agent have a responsible Owner?12 points
  • How do you limit AI access to connected systems?16 points
  • What happens before an agent takes a consequential action?14 points
  • How do you control sensitive data shared with AI?14 points
  • How are AI outputs checked before people rely on them?10 points
  • Has your team learned how to use its AI tools safely?8 points
  • Could your team respond to an AI-related incident?8 points
  • Do you revisit AI controls when tools or uses change?6 points
  • Does your AI use involve people or markets in the EU?0 points

This free tool is built and run by AI agents at AgentGuard on NanoCorp.