Grounding
The property of an AI system's answer being tied to a specific, defined source of information (either loaded organizational context for the AI Briefcase, or verified program content via retrieval for the professor) rather than drawn from unconstrained general training data. Grounding improves relevance and consistency; it does not, by itself, guarantee current or verified accuracy.
Defined in 4 GAGE programs, which carry 5 distinct definitions of it. The wording above is taught in Certified AI Data Governance Professional (CADGP).
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.
The property of an AI system's answer being tied to a specific, defined source of information (either loaded organizational context for the AI Briefcase, or verified program content via retrieval for the professor) rather than drawn from unconstrained general training data. Grounding improves relevance and consistency; it does not, by itself, guarantee current or verified accuracy.
Supplying a model with authoritative source material (usually through retrieval) so that its answer is anchored to that material rather than generated from general training alone. Grounding is the main configuration defense against confident invention, and it is why the context dial is central to reducing hallucination. (see Topic 1.5)
Anchoring an AI's output in supplied source material and instructing it to use only that, shifting the model from guessing to summarizing and measurably reducing fabrication.
Restricting output to material derived from a defined source, with the source indicated. Reduces fabrication and, more usefully, gives a reviewer something to check against.
Whether a genuine source actually supports the specific claim attached to it, and at the strength claimed. The join between claim and source, distinct from provenance.
Where it is taught
The exact lessons this term appears in. The first module of every program is free with a free account.
- Communicating WITH AI: Precision, Structure, and Context-Setting · Practical AI Workflow Design and Prompt Engineering, Certified AI Practitioner: Workplace Foundations
- Source-to-Output Verification Protocols · Critical Thinking and Context Engineering, Certified AI Practitioner: Workplace Foundations
- Your command tools: the Briefcase, the professor, the dossier you will build · Taking the Data Chair, Certified AI Data Governance Professional (CADGP)
- Prompts, context, and why the same model gives different companies different answers · Build Before You Govern, Certified AI Governance Professional (CAIGP)
- AI lifecycle controls · AI Governance, AI Risk, AI Security, and AI Assurance, Certified Enterprise Governance, Risk and Compliance Professional (CEGP)
Where it shows up at work
A definition is where Grounding 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.
Briefs a model like a good manager: context, constraints, examples, verification, and knows when to stop delegating.
One lesson on this page teach it as a primary outcome.
Designs decision rights, committees, intake, approval tiers and escalation so routine uses move and consequential uses get reviewed.
One lesson on this page teach it as a primary outcome.
Tiers vendors by use and impact, requests evidence instead of promises, tests in the customer's context, and plans monitoring and exit.
One lesson 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.
- AI Literacy Leadneeds it at depth
- Chief Compliance Officer, AI compliance editionneeds it at depth
- AI Governance Managerneeds it at depth
- Chief AI Officerneeds it at depth
- Chief Information Security Officer, AI security focusneeds working knowledge
- Chief Risk Officer, AI risk editionneeds working knowledge
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: NIST AI RMF GOVERN; ISO/IEC 42001 Clause 5; NIST AI RMF GOVERN 6 and MANAGE 3; ISO/IEC 42001 Annex A supplier controls.
One interview question about it
Describe a workflow you redesigned around an AI tool. What did the prompt have to contain for the result to be reliable?
Asked of candidates for prompting and workflow design with ai. 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.
- A computer science or software engineering degree carries it as work you can show: Prompting and tool workflows are already in your daily work.
- Privacy and data protection carries it as a claim to back up: Processor and vendor review is routine privacy work.
- Cybersecurity and IT carries it as a claim to back up: Supplier security assessment is routine.
Terms it appears with
Not an alphabetical neighbourhood: these are the terms taught in the same lessons, ranked by how often they appear together.