AI Governance Engineer
GitHub

NIST AI Agent Standards Initiative (2026)

CAISI initiative on interoperable, secure AI agents: identity, authentication, agent security

By Reviewed

From clause to evidenceThe chain from NIST Agents AI Agent Standards Initiative to its evidence record, in 6 steps; the text alternative lists them and the facts under the figure state each in full.Clause: NIST AI Agent Standards Initiative, AI Agent Standards InitiativeClauseNIST AgentsAI AgentStandards…Duty holder: not statedDuty holderNot statedApplies from: no date (Draft or proposed)Applies fromNo dateDraft orproposedArtefact: Agent registry; non-human-identity controls; agent authentication and authorisation; adversarial agent evalsArtefactAgent registryLayers: Layer 03 Evals & Red Teaming as Evidence and Layer 04 Runtime Controls & ObservabilityLayersLayer 03Layer 04Evidence record: the common evidence record (Evidence record v1)Evidence recordEvidence recordv1Same topic elsewhere: no crosswalk topic files this clause yetAs of 2026-09-19 · illustrative, not a claim of conformity From clause to evidenceThe chain from NIST Agents AI Agent Standards Initiative to its evidence record, in 6 steps; the text alternative lists them and the facts under the figure state each in full.Clause: NIST AI Agent Standards Initiative, AI Agent Standards InitiativeClauseNIST Agents · AI Agent Standards…Duty holder: not statedDuty holderNot statedApplies from: no date (Draft or proposed)Applies fromNo date · Draft or proposedArtefact: Agent registry; non-human-identity controls; agent authentication and authorisation; adversarial agent evalsArtefactAgent registryLayers: Layer 03 Evals & Red Teaming as Evidence and Layer 04 Runtime Controls & ObservabilityLayersLayer 03Layer 04Evidence record: the common evidence record (Evidence record v1)Evidence recordEvidence record v1Same topic elsewhere: no crosswalk topicfiles this clause yetAs of 2026-09-19illustrative, not a claim of conformity
From clause to evidence Build the artefact, then file every output it produces as a record that names AIGE-OBL-NIST-AGENTS. Drawn from chapter 08.
Text alternative
  • Clause: NIST Agents, AI Agent Standards Initiative.
  • Duty holder: Not stated.
  • Applies from: No date, Draft or proposed.
  • Artefact: Agent registry.
  • Layers: Layer 03, Layer 04.
  • Evidence record: Evidence record v1.
  • Record schema: Evidence record.
  • No crosswalk topic files this clause yet.
Id
AIGE-OBL-NIST-AGENTS
Instrument
NIST AI Agent Standards Initiative framework
Clause
AI Agent Standards Initiative
Applies from
No date Draft or proposed · Initiative launched 2026-02-17 by NIST CAISI

The artefact that evidences it

Agent registry; non-human-identity controls; agent authentication and authorisation; adversarial agent evals.

What this obligation connects toNIST Agents AI Agent Standards Initiative and its 11 relations: 11 open controls.DirectColour: stack layer of the pattern or control02 Inventory04 Runtime ControlsNIST Agents AI Agent StandardsInitiativeOpen controls (11)AIGE-CTL-AGENT-025 Accountability across hops · DirectAccountability across hopsAIGE-CTL-EVAL-003 Credential Isolation · DirectCredential IsolationAIGE-CTL-AGENT-022 Delegation, never impersonation · DirectDelegation, never impersonationAIGE-CTL-AGENT-002 Its own identity · DirectIts own identityAIGE-CTL-AGENT-020 MCP authorisation (spec 2026-07-28) · DirectMCP authorisationAIGE-CTL-AGENT-032 Plan reviewed before it runs · DirectPlan reviewed before it runs+5 more+5 more: AIGE-CTL-AGENT-021 Replace long-lived secrets with short-lived credentials, AIGE-CTL-AGENT-007 Runtime guardrail on every tool call, AIGE-CTL-AGENT-026 Stopping third-party agents at your boundary, AIGE-CTL-AGENT-037 The person types the secrets, AIGE-CTL-AGENT-036 Typed messages and limited shared memory between agents
What this obligation connects toNIST Agents AI Agent Standards Initiative and its 11 relations: 11 open controls.DirectColour: stack layer of the pattern or control02 Inventory04 Runtime ControlsOpen controls (11)AIGE-CTL-AGENT-025 Accountability across hops · DirectAccountability across hopsAIGE-CTL-EVAL-003 Credential Isolation · DirectCredential IsolationAIGE-CTL-AGENT-022 Delegation, never impersonation · DirectDelegation, never impersonationAIGE-CTL-AGENT-002 Its own identity · DirectIts own identityAIGE-CTL-AGENT-020 MCP authorisation (spec 2026-07-28) · DirectMCP authorisationAIGE-CTL-AGENT-032 Plan reviewed before it runs · DirectPlan reviewed before it runs+5 more+5 more: AIGE-CTL-AGENT-021 Replace long-lived secrets with short-lived credentials, AIGE-CTL-AGENT-007 Runtime guardrail on every tool call, AIGE-CTL-AGENT-026 Stopping third-party agents at your boundary, AIGE-CTL-AGENT-037 The person types the secrets, AIGE-CTL-AGENT-036 Typed messages and limited shared memory between agentsNIST Agents AIAgent StandardsInitiative
What this obligation connects to Solid edges are direct relations, dashed ones looser (a case that cites the same article, a framework whose clauses only touch the topic); patterns and controls take the colour of their stack layer. Select a node to open it; the lists below give every relation in full.
Data table: every relation
Relations of AIGE-OBL-NIST-AGENTS
FamilyItemRelation
Open controlsAIGE-CTL-AGENT-025 Accountability across hopsDirect
Open controlsAIGE-CTL-EVAL-003 Credential IsolationDirect
Open controlsAIGE-CTL-AGENT-022 Delegation, never impersonationDirect
Open controlsAIGE-CTL-AGENT-002 Its own identityDirect
Open controlsAIGE-CTL-AGENT-020 MCP authorisation (spec 2026-07-28)Direct
Open controlsAIGE-CTL-AGENT-032 Plan reviewed before it runsDirect
Open controlsAIGE-CTL-AGENT-021 Replace long-lived secrets with short-lived credentialsDirect
Open controlsAIGE-CTL-AGENT-007 Runtime guardrail on every tool callDirect
Open controlsAIGE-CTL-AGENT-026 Stopping third-party agents at your boundaryDirect
Open controlsAIGE-CTL-AGENT-037 The person types the secretsDirect
Open controlsAIGE-CTL-AGENT-036 Typed messages and limited shared memory between agentsDirect

Patterns that build it

No pattern in the catalogue names this clause on its "Maps to" line yet; the artefact above is the engineering answer.

Open controls that evidence it

Draft controls in the open control profiles that map to this row: each states a requirement and the evidence it must leave behind.

Source

Chapter 08, section NIST AI RMF, checked against its sources on the review date above.

Machine-readable

Cite this obligation

García Aibar, J. (2026). NIST AI Agent Standards Initiative (2026) (AIGE-OBL-NIST-AGENTS). In AI Governance Engineering: The Thesis & Body of Knowledge (v0.5.0). https://doi.org/10.5281/zenodo.22956197. https://aigovernanceengineer.com/obligations/aige-obl-nist-agents. CC BY 4.0

BibTeX

@misc{aige2026obligation,
  author       = {Jorge García Aibar},
  title        = {{NIST AI Agent Standards Initiative (2026) (AIGE-OBL-NIST-AGENTS)}},
  howpublished = {In AI Governance Engineering: The Thesis \& Body of Knowledge},
  year         = {2026},
  version      = {0.5.0},
  doi          = {10.5281/zenodo.22956197},
  url          = {https://aigovernanceengineer.com/obligations/aige-obl-nist-agents},
  note         = {Version 0.5.0}
}