AI Governance Maturity · A0–A4

Where does your regulated AI program sit on the Autonomy Ladder™ — A0 to A4?

Fifteen questions. A deterministic A0–A4 governance-maturity tier for an AI program running under fiduciary, SOC 2, or Reg BI scrutiny — plus the one or two controls holding it a rung below where it should be. Framework and score. No advice, no sales call.

Take the assessment → (5 minutes)See the A0–A4 framework

I authored the Autonomy Ladder™ A0–A4 framework and published it as two open, MIT-licensed pattern libraries — finserv-agent-audit and cre-agent-audit, both Zenodo-archived. This is the same instrument behind autonomy-ladder.io.
DOI 10.5281/zenodo.20434570 · 10.5281/zenodo.20434575

  • A0Informational
  • A1Assisted
  • A2Delegated
  • A3Supervised Autonomous
  • A4Production Autonomous

What you get

01

Your A0–A4 tier

A deterministic governance-maturity tier and 0–60 score, computed the same way every time — no interpolation, no judgment call.

02

The 1–2 controls holding you back

The specific controls scored a rung below, named in plain language — and the exact control that closes each gap.

03

A read for your vertical

The result is framed for the regulated workflow you selected, not a generic maturity grid.

Autonomy Ladder™ — the A0–A4 governance-maturity framework. Open pattern libraries: finserv-agent-audit · cre-agent-audit (Zenodo DOIs). autonomy-ladder.io.
Informational only. This Autonomy Ladder™ self-assessment is a reference framework and score based on self-attested input; no verification is performed. It is not legal, compliance, or regulatory advice, is not a regulator-defensible audit, and forms no attorney-client relationship. Any regulatory characterizations are summaries, not legal conclusions. No reliance should be placed on this result; consult qualified counsel before acting.