Large language model (LLM)

A large language model is a neural network trained on vast amounts of text to predict likely continuations, which in practice lets it draft, summarize, translate, answer questions, and write code. ChatGPT, Claude, Gemini, and the models inside coding and meeting assistants are all LLMs.

For data governance, the key property is that LLMs are hosted services: whatever goes into the prompt travels to the provider's infrastructure, under the terms of the tier the user is signed into. That is why the same model can be low-risk under an enterprise contract and high-risk in a personal account.