# Grounding data, defined with a sourced example

GLOSSARY, read 13 Nov 2026

**The short answer.** The data and context an agent is told to answer from, rather than from what the model remembers. Informatica says its catalog can automatically enrich metadata by applying data classification and association with business glossaries, which it presents as trusted context for AI agents.

## Definition

The data and context an agent is told to answer from, rather than from what the model remembers. Grounding data is only as good as its readiness: a stale or uncertified table makes a confident wrong answer.

## Example from a real tool

**Example.** Informatica says its catalog can automatically enrich metadata by applying data classification and association with business glossaries, which it presents as trusted context for AI agents.

## Related terms

- [Health status](/glossary/health-status/): A short state, such as Healthy or Degraded, computed from checks on an asset like build success, tests and recency.
- [Label withdrawal](/glossary/label-withdrawal/): The removal of a readiness label when the asset no longer qualifies, such as when a test fails or the source loses its label.
- [Governed metric](/glossary/governed-metric/): A business measure, such as net revenue, that has one owned definition used everywhere.

## Where to read more

- [The AI-Ready Data Checklist](/checklist/)
- [All glossary terms](/glossary/)

## Sources

- www.informatica.com/blogs/the-role-of-trusted-context-in-building-ai-agents-how-informatica-idmc-empowers-smarter-data-management.html: https://www.informatica.com/blogs/the-role-of-trusted-context-in-building-ai-agents-how-informatica-idmc-empowers-smarter-data-management.html (read 13 Nov 2026)

Last reviewed 13 Nov 2026. Canonical: https://aireadydatacheck.com/glossary/grounding-data/
