A second assessment · BETA

AI Readiness. Are we ready to switch AI on?

Sabiki Security makes two products. AIRM (Agentic Identity Risk Management) is the ongoing subscription platform that discovers, scores and governs every non-human identity in your Microsoft 365 tenant — start there free with the Sabiki NHI Risk Score™. The AI Readiness Assessment is a different thing: a point-in-time, read-only scan of a single tenant that answers one question with evidence — is it safe to turn on Copilot, and if not yet, exactly what has to change first.

It is in beta. Registration is required before it can be run — by an MSP on a client tenant, or by an end customer on their own. One assessment per customer tenant.

Join the beta waitlist →

What the assessment actually does

Read-only, through the Microsoft Graph API. No agent installed, no data leaves the tenant, nothing is changed.

The verdictA grade, and a plain answer 0–5.0 across four bands — At Risk, Needs Work, Pilot-Ready, AI-Ready — with a stated verdict on whether it is safe to switch AI on today.
112 checksFour pillars Settings (who can let AI in) · Identities (the agents already running) · Data Exposure (how far content travels) · Data Protection (whether it is contained).
Named, not scoredReal objects, not percentages The specific guests, sites, consent grants and applications — each one named, so findings can be acted on the same afternoon.
The coverage ledgerWhat we could see — and what nobody can Every check accounted for: measured, awaiting confirmation, or not assessable on your licence. A control we could not verify is never shown as a pass.
30 / 60 / 90A plan, sequenced by points-per-hour Every finding carries its effort, the grade points it earns back, who owns the fix, whether it is reversible, and how a re-scan verifies it closed.
IndependentNot Microsoft grading its own homework Secure Score tells you a switch is in the wrong position. This names who and what can actually reach your data as a result.

Two ways in. Two different questions.

Both read-only. They answer different things — pick whichever question you actually have.

Free · no registration · anytime Sabiki NHI Risk Score™ — 0–100 “What is already running in my tenant?” AIRM’s own snapshot of the non-human identity estate in minutes. Re-runnable, because tracking it over time is the platform. Start here. Get your Risk Score →
Beta · registration required · one per customer tenant AI Readiness Assessment — 0–5.0 “Are we ready to switch AI on?” The deeper of the two: 112 checks, a verdict and a 30/60/90 plan. A point-in-time answer to a decision you make once, so it is run once per tenant. Join the waitlist →
Why one is gated and the other is not. The Risk Score is AIRM’s front door — open, instant, re-runnable. The Assessment reads deeply across a live tenant and produces a board-grade verdict, so while it is in beta every deployment is registered and approved first, and each customer tenant gets one. Re-scans, drift alerts and score movement are what an AIRM subscription adds.
For managed service providers

Run it for clients. Sell from it.

Registered partners run the Assessment on a client tenant, white-labelled to their own brand. It arrives with its own evidence — and a remediation plan already costed in hours.

Your brandWhite-label the report Your logo, your colours, delivered as your assessment. The client sees your firm doing the work, because you are.
Your services revenueA costed remediation plan Every finding already carries its effort in hours and who owns the fix. That is a scope of work you can quote from on the same call — without writing it first.
Your adoption storyA reason to move now “Not yet ready” with three named criticals is a board conversation. It turns an abstract AI-governance discussion into a dated, evidenced decision.
Your recurring lineThe natural next sale Once a client has seen the number, the question becomes how to keep it moving. That is an AIRM subscription, at partner margin, on a tenant you already manage.
Free AIRM on your own tenant, forever. Every registered partner gets a full AIRM subscription on their own Microsoft 365 tenant at no cost. Run your own house on it before you sell it. Partner programme →
Beta waitlist

Join the AI Readiness beta

The Assessment is in closed beta while we finish the tenant-registration and approval flow. Add yourself to the waitlist and we will come back to you when a place opens — MSPs running it on client tenants and end customers running it on their own are both welcome. Sabiki AIRM itself is generally available today — if you want something you can run right now, start with the free NHI Risk Score.

We will only use this to contact you about the beta. Nothing is scanned until you explicitly grant read-only consent on a tenant, and consent can be revoked at any time. See our privacy policy.

While you wait

The sample report shows exactly what a completed assessment looks like — the grade, the coverage ledger, the named findings and the 30/60/90 plan. Only the tenant is invented.

Download the sample → PDF · 11 pages · no email required

See a sample, then join the list

An assessment on an illustrative tenant. The method, scoring and structure are exactly those of a live run — only the tenant is invented.

PDF · 11 pages · BETA AI Readiness Report — sample Grade and verdict, the coverage ledger, the AI estate by origin, nine ranked findings with remediation math, the 30/60/90 trajectory, and how the assessment differs from Secure Score, SSPM and AI-SPM. Illustrative sample. Download sample →
Register your interest

Open to MSPs running it on client tenants and to end customers running it on their own. Read-only, through the Microsoft Graph API — nothing installed, nothing changed, consent revocable at any time. Each registration is approved before the assessment runs. Places are currently allocated from the waitlist.

Join the beta waitlist → Beta · approved before deployment · one assessment per customer tenant