Least privilege
An established security principle that any actor should be granted only the access it needs to perform its assigned task, for only as long as it needs it. Applied to an agent, it means building up authority from zero and adding back only provably-needed capabilities, rather than granting broad access and trying to remember what to remove.
Defined in 7 GAGE programs, which carry 18 distinct definitions of it. The wording above is taught in Certified AI Governance Professional (CAIGP).
How each discipline defines it
The same term does different work depending on who is using it. These are the definitions as each program teaches them, unedited.
An established security principle that any actor should be granted only the access it needs to perform its assigned task, for only as long as it needs it. Applied to an agent, it means building up authority from zero and adding back only provably-needed capabilities, rather than granting broad access and trying to remember what to remove.
A security principle stating that software should hold only the smallest set of permissions needed to perform its actual function, and nothing more. This topic's three question test (still used, purpose requires it, explainable out loud) applies this principle to a personal device audit.
The principle that an agent (or any account) is granted the minimum access its role requires and no more, reviewed on a schedule and revoked on any anomaly. On the frontier, least-privilege tool scoping is the control that MCP's ease of connection makes most important.
The design principle that an agent should hold only the minimum tools and data access needed to complete its actual task, no more. Named directly in the framework's own "Agent limits" guidance (MGF p.19) and developed in full through this topic's Section 3G method.
Granting an agent only the minimum data and system access its one job requires, and denying everything else, so that even a confused or tricked agent can do only bounded damage. The small-business slice of the deep access-control discipline in Topic 10.7.
Where it is taught
The exact lessons this term appears in. The first module of every program is free with a free account.
- Prompt Injection and Jailbreak Defense · AI Security Fundamentals, Certified AI Practitioner: Workplace Foundations
- Data Poisoning and Model Security · AI Security Fundamentals, Certified AI Practitioner: Workplace Foundations
- Agent-Specific Security Concerns and Mitigation · AI Security Fundamentals, Certified AI Practitioner: Workplace Foundations
- Securing Agent Permissions and Access Boundaries · AI Security Fundamentals, Certified AI Practitioner: Workplace Foundations
- AI Agents for Small Business: Affordable Automation at Scale · Bonus: SMB AI Adoption Path, Certified AI Practitioner: Workplace Foundations
- The Safety Components · The Agent, Deconstructed, Certified Agentic AI Governance Professional (CAAGP)
- Least Privilege for Agents · Bounding by Design, Certified Agentic AI Governance Professional (CAAGP)
- The Enterprise Control Plane · Rollout, Monitoring, and Incident Response, Certified Agentic AI Governance Professional (CAAGP)
- The agent at the door: an AI agent requests database credentials, and you decide · Access and the Keys, Certified AI Data Governance Professional (CADGP)
- Break-glass access: emergencies without permanent holes · Access and the Keys, Certified AI Data Governance Professional (CADGP)
Where it shows up at work
A definition is where Least privilege starts. The lessons above teach it inside the competencies below, which are what job descriptions ask for and what the GAGE credential records. Every link here is drawn from the same careers map that builds a role's readiness check.
Understands prompt injection, data poisoning, model theft, insecure integrations and excessive agent privileges, and the controls that reduce each.
4 lessons on this page teach it as a primary outcome.
Governs AI agents that take actions: tool access, least privilege, interruptibility, cascading actions and accountability for what an agent did.
4 lessons on this page teach it as a primary outcome.
Finds unapproved AI use, sets which tools are approved and what may be pasted, and detects leakage without policing every keystroke.
2 lessons on this page teach it as a primary outcome.
Roles that hire for it
Each role page shows the full competency set, what the seat decides, and a readiness check against your own evidence.
- Chief Information Security Officer, AI security focusneeds it at depth
- AI Security Architectneeds it at depth · 8 open postings this week
- Chief Technology Officer, AI engineering editionneeds it at depth
- AI Governance Engineerneeds it at depth · 7 open postings this week
- Cybersecurity Risk Analystneeds working knowledge
- AI Incident Response Leadneeds it at depth · 1 open posting this week
How job descriptions say it
Postings rarely use the textbook word. These are the phrases the careers map reads as asking for the competencies above.
Where the frameworks place it: OWASP Top 10 for LLM Applications; Google SAIF; NIST AI RMF MEASURE 2; IMDA Model AI Governance Framework for Agentic AI.
One interview question about it
What are the security failure modes specific to AI systems, and which conventional control covers none of them?
Asked of candidates for ai security fundamentals. Answering it in your own words is exactly how a GAGE topic is passed, and the credential records the pass where an employer can check it.
If you come from another field
Some backgrounds already carry this competency, and the readiness check gives credit for it.
- Cybersecurity and IT carries it as a claim to back up: Threat modeling and control design transfer; the AI-specific threats are new.
- A computer science or software engineering degree carries it as a claim to back up: Security fundamentals are a course you took, not yet a control you ran.
- A cybersecurity or information assurance degree carries it as work you can show: Security fundamentals are your coursework and your labs.
Terms it appears with
Not an alphabetical neighbourhood: these are the terms taught in the same lessons, ranked by how often they appear together.