AlphaMethodology

How Alpha rates AI governance.

A transparent, evidence-driven methodology: a five-step governance maturity scale, four standing evidence panels, a six-stage rating process, and a continuous lifecycle that keeps every opinion current.

The governance maturity scale.

Every pillar, and the overall rating, is expressed on a five-step scale from Ad Hoc to Institutional. The scale is anchored to observable behaviors, not aspirations.

  1. AGR 1

    Ad Hoc

    No board-level mandate. AI activity is decentralized, undocumented, and outside enterprise risk.

  2. AGR 2

    Emerging

    Initial policies and a named owner exist. Inventories are partial; controls are inconsistent across business units.

  3. AGR 3

    Defined

    Board-approved AI strategy, model inventory, and risk framework. Controls operate but evidence is not continuously refreshed.

  4. AGR 4

    Integrated

    AI risk is embedded in ERM and audit cycles. Disclosures meet regulator expectations. Agent operations are inventoried and tested.

  5. AGR 5

    Institutional

    Independent oversight, executive accountability, continuous evidence pipelines, and credible engagement with standards and regulators.

Four standing evidence panels.

Alpha's analysts construct every rating from four standing panels. Each data point is dated, sourced, and re-testable - the basis for institutional comparability.

PANEL A

Disclosure Record

Annual reports, proxies, regulator filings, public commitments, and standards participation collected and dated.

PANEL B

Operating Evidence

Board and committee minutes, model and agent registries, change-control logs, third-party AI registers, incident records.

PANEL C

Management Engagement

Structured interviews with named owners across strategy, risk, controls, audit, and oversight.

PANEL D

Independent Signals

Regulator actions, enforcement, litigation, public incidents, and peer benchmarking inputs verified against primary sources.

The rating process, end to end.

Six stages from scope to publication. Analyst work is separated from the rating decision; the issuer's right of reply is procedural, not editorial.

  1. Step 01

    Scope & Notify

    Universe and entity scope confirmed. Rated entities receive a notification pack outlining timeline, evidence panels, and engagement protocol.

  2. Step 02

    Evidence Assembly

    Analysts assemble Panels A-D. Every data point is dated, sourced, and linked. Gaps are recorded explicitly.

  3. Step 03

    Pillar Scoring

    Each of the seven AGR pillars is scored against the maturity scale by the lead analyst using a published rubric.

  4. Step 04

    Committee Review

    An independent Rating Committee reviews the file, challenges the analyst, and votes the rating. Dissents are recorded.

  5. Step 05

    Issuer Right of Reply

    Rated entities receive the draft rationale and may correct factual errors before publication. Opinions are not negotiated.

  6. Step 06

    Publish & Monitor

    The rating, rationale, and outlook are published. The file enters continuous surveillance against trigger events.

Continuous governance lifecycle.

Ratings are not annual snapshots. Alpha runs a continuous loop - Observe, Assess, Engage, Publish - so opinions move with the evidence.

  1. Phase 01

    Observe

    Continuous ingestion of disclosures, filings, incidents, and standards activity across the rated universe.

  2. Phase 02

    Assess

    Trigger events and scheduled refreshes prompt re-scoring of affected pillars, not full re-rating, to keep signal current.

  3. Phase 03

    Engage

    Issuer dialogue on material changes - new agents, restructured oversight, incidents - with documented Q&A on the record.

  4. Phase 04

    Publish

    Outlook changes, watch placements, and rating actions are published with a dated rationale and a clear evidence trail.

Independence, transparency, and the limits of opinion.

What this methodology guarantees

  • - Separation of analyst and rating-committee roles
  • - Dated, sourced, re-testable evidence on every pillar
  • - A published rubric for every maturity step
  • - Continuous surveillance, not annual snapshots

What ratings are not

  • - Investment advice or recommendations on securities
  • - A safety certification of any specific AI model
  • - A substitute for fiduciary, audit, or regulatory judgment
  • - Negotiated outcomes with rated entities

Request the published methodology paper, v2.4.