Free tool · Security of AI use
Know what to fix next.
Answer a short set of questions about your accounts, data and AI use. Get an explained action plan you can work through at your own pace.
Self-reported assessment · Versioned rules · Verified work email required for reports
Register · Verify · Assess
Start with your organisation.
Send verification request from my work emailVerification is manual for this release. Send from the address you registered; an administrator must check the sender and request ID, then reply with a time-limited report access link. This page does not send email automatically. No marketing subscription is created.
Delete this registration and revoke its report access? Local answers are kept until you clear them separately.
Your results · Act
Your action plan
Complete the questions to see your priorities. Your plan will explain the answer behind each recommendation.
Your answers changed. Generate the plan again before relying on these findings.
Use this plan with an AI assistant
This site has not called an AI model. You can copy a briefing to use with your own assistant after reviewing its data-handling terms. No provider receives your answers when you copy.
Marking a task done tracks your progress. It does not change the assessed answer or prove the fix works. Recheck the control, update your answer and generate a fresh plan.
Want help working through it?
Request a guided review or team access by email. Scope, availability and pricing are agreed before any engagement.
Request access or a service →How this check works
The rule set selects relevant questions, puts reported gaps in a suggested order and treats unknowns as verification work. It is not a risk score, certification, penetration test or continuous monitoring service. It may miss issues outside these questions. For an active incident, follow your incident-response process and seek appropriate support.
References: NIST small-business basics · Australian small-business handbook · NIST Generative AI Profile. The questions and triage are Contextual Trust’s interpretation, not an official assessment or endorsement.
Rules ct-security-check/1.0.0 · Sources checked 1 November 2026. Your answers are kept in memory unless you opt into browser storage. Registration details are stored separately; answers are processed by the server only when you request a verified report. This page has no analytics or third-party scripts.