See your own agentic-authority exposure

The Authority Risk Profiler™ turns an inventory of your AI agents, tools, workflows, and data access into one read-only risk profile — so you can see where your authority is ungoverned, and decide. You curate the input; nothing leaves your hands (customer sovereignty).

Authority Risk Index™

Demo data

You learn: How exposed your AI authority surface is, on KYE™'s own rubric. You can: Answer the questions; the score and band update as you go.

Inputs
Results

Score (0–100) 8

  1. Defended

    Your AI authority surface is already at protocol-grade. Replay-proof, structurally meta-governed, banking-defensible. If you want third-party verification, we can run a one-week conformance review (free for already-defended teams as a reference engagement).

  2. Exposed

    You're aware of the risk and have partial controls in place — but the audit chain is not replay-proof, and meta-governance is policy-level only. Recommended: a 2-week PoC-XS to produce a signed Authority Gap™ report on one of your highest-risk workflows.

  3. High-risk

    Material exposure. You have agents acting on customer data or payments without offline-verifiable proof of authorisation. A single uncontrolled action could become a regulator-disclosure event. Recommended: a 4-week PoC-S to integrate one Stack Binding and emit a signed Evidence Pack™ for the workflow — regulator-defensible deliverable.

  4. Critical exposure

    Banking-grade buyers stop reading at this score. Your agents touch regulated workflows AND your authority chain isn't structurally enforced AND you cannot prove what happened offline. This is a board-level risk. Recommended: PoC-M (6–8 weeks, two Stack Bindings, regulator-replayable verification deck) OR direct Pilot.

How this widget is built
Demonstrates
KYE™ quizzes
Components
  • Quiz manifest (scores per answer)
  • Scoring bands
Flow
Answers → score → band → recommendation. Architecture diagram
Used on

Edit the inventory below and your profile recomputes live, in your browser. The same deterministic scorer KYE Protocol™ runs server-side runs here — nothing is sent anywhere. It loads with a sample bank estate so you can see it working; replace it with your own.

Your inventory

A customer-curated kye.risk_profile_input.v1 — include only what you choose (agents · tools · workflows · data access · the controls you already have). It is read-only and stays in your browser.

Loading a sample profile…

Your posture at a glance

One number, honestly derived. The tier maps to the EU AI Act risk floor.

Where you're governed — across the lifecycle

Every consequential action travels the agentic-governance lifecycle. A stage with no declared control is ungoverned authority. Green is governed; red is a gap.

Your estate — clutter, overlaps, concentrations, bottlenecks, painpoints

This is the structural picture of your own systems, derived from what you included. It helps you understand your estate and act — a single overloaded tool, two redundant agents, a concentration of high-risk authority.

Top exposures & what to add

The sharpest individual risks, ranked — and the controls KYE™ would govern to close each gap, additive to your stack.

Canonical KYE™ surfaces referenced on this page: KYE Protocol™ · Systemic Authority Risk™.