The AIQ™ Score A New Standard for AI Quality
The first independent, quantitative rating of enterprise AI governance — a 0-to-200 score built on 250 proprietary data points across five dimensions of AI governance maturity, designed for capital partners, boards, insurers, and regulators.
Initializing...
Hover over a framework to see its impact on our scoring methodology
Methodology
What We Measure
The AIQ™ Score quantifies the maturity and integrity of an organization’s AI systems across five dimensions, using 250 proprietary data points — 50 per dimension — drawn from disclosed data, proprietary surveys, and AI-assisted verification.
Strategic Alignment
20%Executive commitment and business integration
Indicators include CAIO presence, AI investment levels, and strategy disclosure.
Oversight & Accountability
30%Governance structures, board oversight, and policy maturity
Indicators include board reporting cadence, audit frequency, and policy documentation.
Technical Robustness
25%Technical controls, model quality, and operational integrity
Indicators include model validation, security testing, bias audits, and MLOps practices.
Responsible AI & Compliance
15%Fairness, transparency, and regulatory compliance
Indicators include bias mitigation, explainability, and regulatory alignment.
Adaptability & Education
10%Continuous improvement and innovation capacity
Indicators include incident learning, retraining frequency, and feedback loop maturity.
Weighted Methodology
Each dimension is weighted to reflect its relative impact on overall AI governance quality. Oversight & Accountability carries the highest weight at 30% because governance failures are the leading driver of AI-related loss events and regulatory enforcement. Technical Robustness follows at 25%, reflecting the operational risk of poorly built systems. Strategic Alignment (20%), Responsible AI & Compliance (15%), and Adaptability & Education (10%) complete the model. Weights are informed by risk exposure analysis, regulatory emphasis, and actuarial data — and are reviewed annually.
The 0–200 Scale
The AIQ™ Score uses a 0–200 scale modeled on two proven precedents: the standard IQ scale used in psychometric assessment, where 100 represents the population mean, and the Ocean Tomo Patent Ratings® scale used to rate patent quality for the NYSE-listed OT300® Index. The 200-point range provides the granularity to differentiate governance maturity.
Nascent
0–40
Minimal formal AI governance; ad hoc practices
Considering
40–70
Beginning to explore AI governance frameworks
Developing
70–100
Foundational governance in place; significant gaps remain
Established
100–130
Mature governance with documented practices and oversight
Advanced
130–160
Advanced governance exceeding baseline requirements
Leading
160–200
Exemplary governance; industry-leading practices across all dimensions
Organizations completing the full 12-month AIQ™ Assessment and achieving verified AIQ™ Scores of 115 or above — within the Established band — earn AIQA Certification, the outcome credential of the AIQ™ Assessment program.
Framework Alignment
The AIQ™ methodology is continuously updated to reflect the latest releases from global AI governance frameworks and the evolving consensus on industry best practices—keeping the rating aligned with the current regulatory and standards landscape rather than any single point-in-time snapshot.
Primary anchors
Additional cross-jurisdiction anchors
These reinforce — they do not replace — the core five.
Provider vs Deployer Assessment
AI governance obligations are not symmetric — building an AI system carries different responsibilities than deploying one built by others. The AIQ™ score adapts its emphasis to the organization’s role, so that scores reflect the obligations that actually apply, rather than penalizing deployers for the absence of provider artifacts or providers for the absence of operational use logs.
Building & Releasing AI Systems
For organizations that design, develop, or substantially modify AI systems, the AIQ™ score weights Technical Robustness and Oversight & Accountability most heavily. Evidence centers on the artifacts that demonstrate a system’s fitness for release: Annex IV technical documentation, model cards, training-data provenance, red-team results, conformity assessments, and post-market monitoring plans. The assessment anchors to provider obligations under EU AI Act Articles 16–22, with equivalent NIST AI RMF and ISO/IEC 42001 controls applied in parallel. Organizations placing high-risk AI systems on the EU market activate the Provider Overlay, which scores the EU-specific provider regime in depth and qualifies them for the differentiated AIQA Certified — High-Risk AI Provider designation.
Integrating AI into Operations
For organizations that put AI systems to work — whether built in-house or acquired from a vendor — the AIQ™ score weights Oversight & Accountability and Strategic Alignment most heavily. Evidence centers on how the organization governs use: AI system inventories, use-case classifications, human-oversight logs, incident and remediation records, DPIAs, and vendor risk reviews. The assessment anchors to deployer obligations under EU AI Act Article 26, GDPR provisions on automated decision-making, and equivalent NIST AI RMF and ISO/IEC 42001 controls. Deployer-only organizations are scored against the base 250-data-point methodology without the Provider Overlay.
Validation & Data Integrity
AIQ™ Scores undergo cross-validation, inter-rater reliability testing, and statistical correlation analysis to ensure consistency and defensibility. AIQA’s methodology supports privacy-preserving assessment architecture, allowing organizations to participate in governance evaluation without exposing proprietary model details, architectures, or training data to external evaluators.