Sylvan Assurance AI Security · Free Edition ← Free assessments
Free AI security maturity assessment

How mature is your AI security programme?

Twenty-eight plain-English checks across the eight working parts of AI security — AI inventory and shadow AI, vendor and model supply chain, prompt injection and input handling, data protection in AI pipelines, access control and agentic systems, monitoring and drift, incident readiness, and governance. You get a scored snapshot with plain-English next steps. About 12 minutes. No email required.

Plain-English security — without us ever seeing your answers. Everything stays in your browser. Nothing is transmitted or tracked.

Don't take our word for it — verify in 30 seconds.
  1. Right-click this page and choose Inspect, then open the Network tab.
  2. Start the assessment and answer a few questions.
  3. Watch the list: nothing leaves the page while you answer — your answers stay on your device.
What does this assessment cover?

The eight working parts of AI security. The assessment starts where the books start: AI inventory and shadow AI (knowing what is actually in use, including the tools staff signed up for on their own) and the vendor and model supply chain. It then covers prompt injection and input handling, data protection in AI pipelines, and access control and agentic systems. The operations half follows: monitoring and drift, incident readiness, and governance and reporting.

What you get. A weighted maturity score with an area-by-area breakdown and plain-English priority actions, starting from your weakest area. One honest note from the books: the right level of control is the one proportionate to how much AI your organisation actually runs — a small estate with three tools can be exactly where it should be.

What it isn't. This is general guidance. It is not a professional audit, not a penetration test, and not legal advice. Every recommendation is optional. Following it reduces common risks but does not guarantee any outcome. Responsibility for your programme remains with you.

Start here — the free edition

Version 1 · Updated 2026-07-21

For anyone who wants to see where they stand with AI security before deciding whether to go further.

When you have worked through this, you will have: your free maturity result, and the five first fixes worth doing before anything else. Download this as a PDF.

The one guide
The First 5 AI Security Fixes
PDF · 20 min to read · Whoever has just been handed AI security
Five sequenced starter actions — take an inventory, read the training clause, write the never-paste rule, cut back access, decide the kill switch — with roughly a week allowed for each.
Reach for it when: First. It is the free starter, and the whole shape of the paid toolkit in miniature.
You finish with: Five named actions with dates against them.
The interactive assessment — a short questionnaire that scores where you stand and points at the gaps.

When you outgrow the free tier, the Solo toolkit turns these five fixes into a worked programme — a written AI-tool register, a one-page acceptable-use policy, a first threat model, and more — with a real Start here catalog to sequence them. See the Solo, Team and Enterprise editions →

When you're ready for the full toolkit

The full edition gives you the complete toolkit.

Everything behind this free assessment — the working documents, templates, runbooks, and depth to put it into practice. Three editions to fit how you work.

See the full toolkit & pricing →

One-time purchase · files you own forever · 30-day money-back guarantee.