What is it?
Grounding
Tying a model's output to verified facts or real sources rather than letting it invent things. Memory and retrieval both provide grounding: the agent answers based on what it actually knows about the user, the project, or the retrieved documents. Good grounding reduces hallucination.
From the AI agent glossary
As the Planning and Reasoning glossary puts it
Grounding means anchoring a model's statements to real, retrieved information rather than letting it rely on memory or guesswork. When a ReAct agent searches for a fact and then reasons from the result, its conclusion is grounded in that retrieved evidence. Grounding is the main defense against hallucination: it keeps the model's reasoning tethered to things that are actually true.
Where Grounding is explained
- ReAct, Reasoning and Acting TogetherPlanning and Reasoning · A worked example
- Context Engineering and MemoryArchitecture and System Design · Reference, don't embed
- Retrieval: RAG vs Tools vs Long ContextArchitecture and System Design · Why tools are often the right default for document agents
- Observability and EvaluationArchitecture and System Design · From observability to evaluation
- Domain-Specific and Regulated-Industry AgentsArchitecture and System Design · Evaluation with domain rigor