Glossary · AI and language-model infrastructure
Large language model
A machine-learning model trained to process and generate language by predicting likely sequences from context.
Also called: LLM, language model
Definition
A large language model can summarize, classify, generate, transform, or reason over text and other supported inputs. Its response is influenced by training, prompt, available context, tools, sampling, and system constraints.
Why it matters
Language models can support useful interfaces and workflows, but outputs need validation proportional to the business, privacy, and operational risk.
Common misunderstanding
A fluent response is not proof that the model retrieved a fact, followed current policy, or reached a correct conclusion.
Related terms
Related system family: Operations & Automation
Sources
- Artificial intelligence (opens in a new tab) · NIST Computer Security Resource Center
Last reviewed July 16, 2026