Skip to main content

CAI Governance Readiness Model

Purpose​

This model determines whether governance exists as a precondition of execution.

It does not describe governance programs or organizational maturity.

It evaluates whether governance structurally determines if execution can occur.

Core Test​

At every level, evaluate one condition:

Can a known invalid action execute?

  • If YES, the system fails the level
  • If NO, the system passes the level

Model Structure​

The model progresses through six levels.

Each level introduces a stricter condition on execution.

Progression is not cumulative description.

It is cumulative elimination of execution paths for invalid actions.

System Boundary Definition​

The system under evaluation must be explicitly defined.

The boundary must include:

  • All components required for the action to occur
  • All control points where the action could be permitted or blocked
  • All alternate execution paths available to the actor

If any execution path exists outside the defined boundary that allows the action:

The system fails classification.

Boundary Completeness Requirement​

The defined system boundary must account for all execution dependencies.

This includes:

  • External systems
  • Third-party services
  • APIs and integrations
  • Indirect execution paths
  • Shadow or undocumented infrastructure

The evaluator must demonstrate:

  • That no execution path exists outside the defined boundary
  • That all components capable of enabling the action are included

If any external or unaccounted path exists:

The system fails classification.

Path Completeness Constraint​

Classification applies only to known and observable execution paths.

The evaluator must demonstrate:

  • That all execution paths have been enumerated
  • That no undocumented or hidden paths are likely to exist

If path completeness cannot be established:

The system fails classification.

Admissibility Integrity Requirement​

Admissibility must be enforced through non-bypassable constraints.

Admissibility does not include:

  • Spoofable tokens
  • Reversible checks
  • Identity claims without verification
  • Logic that can be externally manipulated

If admissibility can be bypassed without violating system constraints:

The system fails classification.

Level 0 - No Binding​

Test​

Can an invalid action execute?

Result​

YES

Failure Reason​

No system-level representation of governance exists.

Level 1 - Post-Execution Awareness​

Test​

Can an invalid action execute before detection?

Result​

YES

Failure Reason​

Detection occurs after execution.

Level 2 - Pre-Execution Evaluation Without Authority​

Test​

Can an invalid action execute even after being identified as invalid?

Result​

YES

Failure Reason​

Evaluation exists but does not determine execution.

Level 3 - Partial Execution Gating​

Test​

Are there any execution paths where invalid actions can still occur?

Result​

YES

Failure Reason​

Governance is enforced inconsistently across execution paths.

Level 4 - Complete Execution Binding​

Test​

Can any invalid action execute across all defined execution paths?

Result​

NO

Failure Reason (if failed)​

At least one execution path bypasses admissibility.

Level 5 - Constraint Invariance During Execution​

Test​

Can a valid action transition into an invalid state during execution?

Result​

NO

Failure Reason (if failed)​

Constraints do not persist across state transitions.

Constraint Invariance Clarification​

Constraint invariance requires that constraints remain valid throughout execution.

This includes:

  • Asynchronous processes
  • Concurrent state changes
  • Distributed system updates

If execution can proceed under conditions that invalidate constraints at any point:

The system fails Level 5.

Required Inputs​

Every evaluation must specify:

  • Defined invalid action
  • Execution path
  • Observed system behavior

If any of these are missing, the system defaults to failure.

Required Output​

Every evaluation must produce:

Level Achieved: [0-5] First Failing Level: [0-5]

Failure Condition: [Exact test failure]

Bypass Path: [Where execution escaped governance, if applicable]

Interpretation Rules​

  • If ambiguity exists, default to failure
  • If evaluators disagree, the system fails the level
  • If no invalid action can be defined, the system is Level 0

Core Principle​

The defining transition occurs between Level 2 and Level 3.

At Level 2, governance evaluates execution.

At Level 3 and above, governance determines whether execution exists.

Structural Function​

This model replaces descriptive maturity assessment with executable verification.

It does not ask what governance looks like.

It asks whether invalid actions are structurally impossible.

If invalid actions can occur, governance does not bind execution.

If invalid actions cannot occur, governance exists as a precondition of execution.