Skip to main content
The Alpha Standard is the complete methodology behind Alpha Governance Ratings. It provides independent, evidence-based assessments of how companies govern artificial intelligence. The ratings answer one question: Is this company governing AI in a way that protects stakeholders and enables sustainable competitive advantage? We measure governance effectiveness, not AI capability or strategy quality. We measure outcomes and stress responses, not just governance artifacts.

​
Evidence hierarchy

Not all governance evidence is equal. Alpha applies a three-tier evidence model:
Evidence tierExampleCredit
ArtifactPublished AI ethics policyPartial
ApplicationEvidence the policy was applied in a real decisionFull
Stress testEvidence governance held under competitive pressureHighest
A company with a comprehensive AI ethics policy (artifact) that was never invoked when a model showed bias in production has weaker governance than a company with a simpler policy that demonstrably triggered a model rollback.

​
Four phases

1

Phase 1: Alpha Materiality Profile

Assess how material AI governance is for this specific company. Four dimensions: operational dependency, strategic centrality, risk exposure, and stakeholder impact. This determines the weight of the overall rating in the company’s risk profile.
2

Phase 2: The Alpha 10

Score governance quality across 10 dimensions, weighted by company profile (Developer, Deployer, or Hybrid). Each dimension is scored 0-100 with specific evidence requirements.
3

Phase 3: Alpha Integrity Screen

Apply negative screening across five concern categories. Critical failures impose grade ceilings. A company cannot achieve an A-tier rating while maintaining a pattern of governance violations.
4

Phase 4: Alpha Rating Committee

Final determination. Assign the letter grade, trajectory overlay (Advancing / Stable / Declining), and watch status (Positive / Negative / None). Peer comparison is included in every rating.

​
Rating profiles

A foundation model developer and a bank deploying AI face fundamentally different governance challenges. Alpha uses three profiles within the same dimensional framework:
ProfileApplied toHeavier weight on
DeveloperCompanies that build, train, and distribute AI modelsSafety, Data Ethics, Cybersecurity
DeployerCompanies that integrate, deploy, and operate AI systemsGovernance Infrastructure, Fairness, Workforce, Third-Party
HybridCompanies that both build and deployBlended weights
The rating report always discloses which profile was applied and why.

Full dimension weights

See the complete weight table for all 10 dimensions across Developer, Deployer, and Hybrid profiles.

​
Committee Accountability Mapping

The 10 dimensions are not independent scoring lanes. They cluster into five committee accountability zones, reflecting how governance actually works at the board level.
Committee zoneDimensionsZone weight
Full BoardD1: Leadership and Accountability15%
Risk CommitteeD2: AI Safety, D4: Data Ethics, D9: Cybersecurity30%
Audit CommitteeD3: Transparency, D8: Regulatory, D10: Third-Party25%
Compensation/HRD6: Fairness, D7: Workforce/Societal15%
Nominating/GovernanceD5: Governance Infrastructure15%
High variance between dimensions within the same committee zone flags a structural governance gap. If Transparency scores 80 but Regulatory Compliance scores 40, the Audit Committee is not providing coherent oversight.