NIST AI RMF crosswalk: the four functions, mapped
If your AI risk program needs to implement the four NIST AI RMF functions - GOVERN, MAP, MEASURE, and MANAGE - here is how each maps to the AGLedger record, the Signed Statement chain, the dispute path, and the Settlement Signal™ surface.
The NIST AI RMF is a voluntary framework and there is no certification against it. This page describes how AGLedger's controls map to its functions; it is not a certification and does not confer one.
The crosswalk is a capability mapping. AGLedger provides the evidence pattern; your program provides the risk methodology, the tolerances, and the response decisions.
If your AI program needs to evidence NIST AI RMF MEASURE controls, the crosswalk below applies. The same evidence pattern - signed records, hash-chained, append-only - also supports non-AI automated work in any control family that expects tamper-evident audit trails for RPA, CI pipelines, and microservice calls. AGLedger is software you self-host; the regulations are AI-framed, the underlying evidence pattern is not.
The crosswalk at a glance
| Function | AGLedger evidences | Your program owns |
|---|---|---|
| GOVERN | Who held which authority, and every oversight action, on a signed chain (record → completion → verdict) | Policies, risk tolerances, who is responsible |
| MAP | Risk level and domain classification per record; each operation tied to its risk context | The risk assessment itself; categories and thresholds |
| MEASURE | Tolerance bands on criteria, signed timestamps on every transition, queryable acceptance and rejection rates | Measurement criteria; interpreting results |
| MANAGE | Dispute and remediation states on the chain; Settlement Signal webhooks routing outcomes; full export for third-party audit | Risk response strategy; resource allocation |
Each function is expanded below. For NIST's own mappings to ISO/IEC 42001 and the EU AI Act, see the official AI RMF crosswalks.
GOVERN
Establish and maintain policies, processes, and accountability structures for AI risk management.
AGLedger provides
- Structured accountability chain for every automated operation (record → completion → verdict)
- Every record names its principal and its performer; the designated overseer, their authority scope, and the designation date are recorded as structured metadata
- Append-only audit vault records every policy decision and oversight action
- Per-record compliance attestation records linked to the audit chain
You own
- Defining governance policies and risk tolerances
- Designating responsible individuals and their authority
- Establishing organizational AI risk management strategy
MAP
Identify, categorize, and document AI risks in context.
AGLedger provides
- Risk level and domain classification per record (maps to Annex III categories)
- Structured records linking each automated operation to its risk context
- Federation carries risk mapping between Servers, most often several Servers inside one company; cross-company federation is the rarer case
- Custom schemas allow domain-specific risk categorization
You own
- Performing the risk assessment
- Determining risk categories and thresholds
- Mapping AI systems to organizational context
MEASURE
Analyze, assess, and track AI risks and impacts.
AGLedger provides
- Tolerance bands enforce numeric bounds on record criteria
- Timeliness evidence - every state transition timestamped and signed
- Agent drift: what each agent did in the current window, the window before, and the difference, with no score or threshold
- Audit vault queryable for cross-record analysis of acceptance, rejection, and revision rates
You own
- Defining measurement criteria and acceptable thresholds
- Interpreting measurement results
- Deciding what corrective action to take
MANAGE
Allocate resources and implement plans to respond to AI risks.
AGLedger provides
- Dispute resolution: a signed evidence window, then an outcome the principal renders
- Remediation states and revision workflow for corrective actions
- The Settlement Signal API emits SETTLE, HOLD, and RELEASE to route outcomes to downstream systems
- Full chain exportable for regulatory submission and third-party audit
You own
- Resource allocation decisions
- Risk response strategy and implementation
- Ongoing monitoring program design
The AI RMF is voluntary, but its GOVERN, MAP, MEASURE, and MANAGE functions all assume durable evidence of how your AI systems actually behave in production. AGLedger provides that evidence pattern - signed, hash-chained, exportable for third-party audit; your risk-management program provides the policy and process around it.
Frequently asked
- How does AGLedger map to the NIST AI RMF GOVERN function?
- GOVERN establishes and maintains policies, processes, and accountability structures for AI risk management. AGLedger provides a structured accountability chain for every automated operation (record to completion to verdict), a principal and a performer named on every record with the designated overseer and their authority scope held as structured metadata, an append-only audit vault that records every policy decision and oversight action, and per-record compliance attestation records linked to the audit chain. You own defining governance policies and risk tolerances, designating responsible individuals, and establishing the organizational AI risk management strategy.
- How does AGLedger support the MAP function?
- MAP identifies, categorizes, and documents AI risks in context. AGLedger provides risk level and domain classification per record (mapping to Annex III categories), structured records linking each automated operation to its risk context, federation that carries risk mapping between Servers (most often several Servers inside one company, with cross-company federation the rarer case), and custom schemas for domain-specific risk categorization. You own performing the risk assessment, determining risk categories and thresholds, and mapping AI systems to organizational context.
- How does AGLedger support the MEASURE function?
- MEASURE analyzes, assesses, and tracks AI risks and impacts. AGLedger provides tolerance bands that enforce numeric bounds on record criteria, timeliness evidence with every state transition timestamped and signed, agent drift (per-window counts of what each agent did and the change between windows, with no score or threshold), and an audit vault queryable for cross-record analysis of acceptance, rejection, and revision rates. You own defining measurement criteria, interpreting results, and deciding what corrective action to take.
- How does AGLedger support the MANAGE function?
- MANAGE allocates resources and implements plans to respond to AI risks. AGLedger provides dispute resolution (a signed evidence window, then an outcome the principal renders), remediation states and a revision workflow for corrective actions, Settlement Signal webhooks that route SETTLE, HOLD, and RELEASE outcomes to downstream systems, and a full chain exportable for regulatory submission and third-party audit. You own resource allocation decisions, risk response strategy, and ongoing monitoring program design.
Primary sources
- NIST AI 100-1: Artificial Intelligence Risk Management Framework (AI RMF 1.0) - the framework itself, and the source of the four function definitions quoted above
- NIST AI RMF Playbook - the suggested actions per subcategory, which is where the evidence expectations get concrete
- NIST AI RMF crosswalks - NIST's own mappings to ISO/IEC 42001, the EU AI Act, and other frameworks