Skip to main content
    CALCULATORiQ™

    CALCULATORiQ Research · Global AI Assurance

    AI CAPABILITY AND RISK MATRIX

    Place a frontier AI system on the five tier capability ladder, score it across four systemic risk vectors, benchmark assurance readiness against major frameworks and size a remediation capital reserve.

    Read the Global AI Assurance Constitution

    System profile

    Training compute1e26 FLOP

    Order of magnitude of training FLOP.

    Unsupervised task horizon8 h

    Hours the system can work without human review.

    Agentic tool access55

    Breadth of tools, APIs and actions available.

    Self-improvement coefficient15

    Ability to modify its own code, weights or training.

    Critical infrastructure exposure40

    Grid, water, telecoms, health and transport integration.

    Financial execution authority30

    Ability to move funds or execute trades.

    Assurance control maturity50

    Evaluations, audit trails, kill switches, incident reporting.

    Deployed capital$5.0bn

    Capital committed to the deployment.

    Capability tier

    Tier 4 · Frontier systems

    Capability index 61 / 100

    Composite residual risk

    33

    After assurance controls

    Risk vectors

    Systemic cyber contagion40
    Critical infrastructure vulnerability33
    Autonomous financial execution27
    Sovereign containment29

    Assurance readiness by framework

    EU AI Act58%
    US NIST AI RMF63%
    G7 Hiroshima Process64%
    BRICS+ multilateral norms66%

    Remediation capital reserve

    $146m

    2.93% of deployed capital, set aside for incident remediation, consistent with a Frontier Assurance and Remediation Fund contribution model.

    Illustrative scoring model for governance research. Weightings are editorial assumptions and do not represent a regulatory determination, legal advice or an investment recommendation.