Agentic Governance Benchmark (AGB)
A standardized framework for measuring runtime governance enforcement in autonomous AI systems.
Use the free browser-based self-assessment to score six dimensions against the published rubric and download a PDF report. Your inputs produce a weighted composite from 0 to 100; this tool does not test your system or independently verify your evidence.
What AGB Measures
AGB scores runtime governance enforcement across six weighted dimensions. The composite score is the sum of each dimension's score multiplied by its weight. The framework is vendor-agnostic and scores individual systems, not governance across organizational boundaries.
- Dimension
- D1 · Policy Determinism
- Weight
- 25%
- What it measures
- Same input produces the same governance outcome every time. Enforcement is binary pass/fail with no probabilistic thresholds and no model-dependent variance.
- Dimension
- D2 · Enforcement Latency
- Weight
- 20%
- What it measures
- Violations are caught before the action executes. Pre-execution enforcement means the action cannot proceed without clearance. Post-execution logging is incident response, not governance.
- Dimension
- D3 · Receipt Provenance
- Weight
- 20%
- What it measures
- Every governance decision produces a cryptographically signed receipt tying intent, policy version, decision outcome, and execution boundary together. Receipts are Merkle-chained into a tamper-evident ledger.
- Dimension
- D4 · Scope Containment
- Weight
- 15%
- What it measures
- Provable evidence the agent never exceeded its authorized action space. Scope containment means verifiable boundaries, not absence of detected violations.
- Dimension
- D5 · Jurisdictional Enforcement
- Weight
- 10%
- What it measures
- The correct regulatory framework applies automatically per request based on data residency, user location, and operation type. No manual jurisdiction selection.
- Dimension
- D6 · Override Integrity
- Weight
- 10%
- What it measures
- Human overrides pass through the same enforcement layer as autonomous actions. Every override is policy-checked, receipted, and auditable.
| Dimension | Weight | What it measures |
|---|---|---|
| D1 · Policy Determinism | 25% | Same input produces the same governance outcome every time. Enforcement is binary pass/fail with no probabilistic thresholds and no model-dependent variance. |
| D2 · Enforcement Latency | 20% | Violations are caught before the action executes. Pre-execution enforcement means the action cannot proceed without clearance. Post-execution logging is incident response, not governance. |
| D3 · Receipt Provenance | 20% | Every governance decision produces a cryptographically signed receipt tying intent, policy version, decision outcome, and execution boundary together. Receipts are Merkle-chained into a tamper-evident ledger. |
| D4 · Scope Containment | 15% | Provable evidence the agent never exceeded its authorized action space. Scope containment means verifiable boundaries, not absence of detected violations. |
| D5 · Jurisdictional Enforcement | 10% | The correct regulatory framework applies automatically per request based on data residency, user location, and operation type. No manual jurisdiction selection. |
| D6 · Override Integrity | 10% | Human overrides pass through the same enforcement layer as autonomous actions. Every override is policy-checked, receipted, and auditable. |
Maturity Tiers
AGB maps composite scores to five maturity tiers:
- Tier
- Ungoverned
- Score range
- 0 - 14
- Description
- No runtime enforcement. Compliance is aspirational or nonexistent.
- Tier
- Reactive
- Score range
- 15 - 39
- Description
- Periodic audit-based compliance. Governance is procedural, not structural.
- Tier
- Monitored
- Score range
- 40 - 69
- Description
- Post-execution detection with structured logging. Violations caught but not always prevented.
- Tier
- Governed
- Score range
- 70 - 89
- Description
- Strong enforcement with minor gaps. Pre-execution on critical paths. Structured audit trail.
- Tier
- Sovereign
- Score range
- 90 - 100
- Description
- Deterministic enforcement with cryptographic provenance. Governance operates as infrastructure. Real-time compliance provable on demand.
| Tier | Score range | Description |
|---|---|---|
| Ungoverned | 0 - 14 | No runtime enforcement. Compliance is aspirational or nonexistent. |
| Reactive | 15 - 39 | Periodic audit-based compliance. Governance is procedural, not structural. |
| Monitored | 40 - 69 | Post-execution detection with structured logging. Violations caught but not always prevented. |
| Governed | 70 - 89 | Strong enforcement with minor gaps. Pre-execution on critical paths. Structured audit trail. |
| Sovereign | 90 - 100 | Deterministic enforcement with cryptographic provenance. Governance operates as infrastructure. Real-time compliance provable on demand. |
How to score your system
The assessment takes about five minutes. Score what you could demonstrate to an auditor today, not what is on the roadmap. Nothing you enter leaves your browser.
Set each of the six dimensions
Select the capability level you can demonstrate. If your evidence supports a score between levels, open the optional exact-score field. Intermediate scores represent partial attainment of the next level.
Score the evidence, not the intent
Pick the level you could prove to an auditor today. A control that exists but produces no inspectable record scores where the evidence sits, not where the design does.
Read the composite as you go
Each dimension is multiplied by its weight and summed into a score out of 100. The card updates live with your maturity tier and a radar chart of where the gaps are.
Export or share the result
Download a branded PDF readiness report with dimension scores and suggested capabilities to investigate, or copy a plain-text summary for a ticket or email.
Terms used in this assessment
- Deterministic
- Same input always produces the same governance outcome.
- Pre-execution gate
- Clearance is required before an action runs, not after.
- Receipt
- Tamper-evident proof of what happened and under which policy.
- Scope
- The boundary of tools, data, and permissions an agent may use.
- Jurisdiction
- Which regulatory rules apply to a given request.
- Override
- A human exception that is still policy-checked and audited.
Maturity Tiers
Interpreting the result
AGB measures governance capability, not governance outcomes. Even a Sovereign system can make bad decisions if its policies are poorly designed. The paper identifies independent assessment protocol development and empirical validation of the weights as further work.
- Measurement area
- Dimension scores
- Supported interpretation
- Evidence supports the selected capability level.
- Unsupported interpretation
- The score reflects planned capability or unsupported assumptions.
- Measurement area
- Composite and tier
- Supported interpretation
- The six weighted scores determine a maturity tier.
- Unsupported interpretation
- A tier is presented as certification or proof of regulatory compliance.
- Measurement area
- Governance outcomes
- Supported interpretation
- Policy quality and operational outcomes are evaluated separately.
- Unsupported interpretation
- A high enforcement score is assumed to guarantee good decisions.
| Measurement area | Supported interpretation | Unsupported interpretation |
|---|---|---|
| Dimension scores | Evidence supports the selected capability level. | The score reflects planned capability or unsupported assumptions. |
| Composite and tier | The six weighted scores determine a maturity tier. | A tier is presented as certification or proof of regulatory compliance. |
| Governance outcomes | Policy quality and operational outcomes are evaluated separately. | A high enforcement score is assumed to guarantee good decisions. |
Companion: Override Health Benchmark
AGB measures runtime governance, including human override integrity in D6. The Override Health Benchmark, or OHB, is a separate companion framework for human governance: whether oversight remains meaningful, timely, and durable when autonomous systems push toward action.
OHB extends the discussion of human oversight beyond AGB’s override-integrity dimension. It is separate from the six-dimension AGB composite described in the white paper.