AG-A-02Level 1 · Signal
Agent holds broad or standing permissions
- Access
- Neutral
- Level 1 · Signal
An Access signal for an AI agent from identity governance (IGA, agent scope inventory): the agent's scopes are wider than its task or carry no expiry.
Classification
- Risk domain One of 9 areas of human risk, from social engineering to readiness and response. Every signal belongs to one.
- Evidence level
- Maturity practices One capability a program performs, scored on its own: See, Understand, Act, Prove, Govern and Culture.
- Leads to The kind of breach the evidence beneath it points toward, each anchored to a Verizon DBIR or VERIS pattern or an FBI IC3 category.
- NIST CSF 2.0
-
- PR.AA-05
- MITRE ATT&CK
-
- T1078.004
The definition, in detail
Definition
That reach is the blast radius if its credentials leak or its owner is compromised.
Example
An invoice-processing agent holds write access to every finance table and a token that never expires, though its task touches one queue.
How it is measured
- Source
- IGA, PAM & entitlements
- Subject
- AI agent
- Event
- Agent scope inventory
- Risk domain
- AI agents
- Provenance
- Framework research
How it is used
Programs count "Agent holds broad or standing permissions" per ai agent from IGA, PAM & entitlements and read it beside the other AI agents signals; it feeds 1 cohort, 0 sequences and 2 breach scenarios.
A "agent holds broad or standing permissions" record is evidence of what the person could expose (the access side of the model). It describes reach; it records no action by the person and no attack under way.
Technical mappings
Framework mappings; validate for your use case.
Continue exploring
"Agent holds broad or standing permissions" (AG-A-02), Human Risk Management Framework v2.0.0, Living Security, CC BY 4.0. https://www.humanriskmanagement.com/taxonomy/signals/ag-a-02-agent-holds-broad-or-standing-permissions
Part of the open HRM Taxonomy, CC BY 4.0. Download the taxonomy as JSON · Version history. This page describes one signal.
The initial assessment
Find your level.Then fix what matters.
18 questions, about 4 minutes and nothing to type before the result: your level, a score for each practice and three first playbooks, each Easy, Medium or Hard.