Glossary

Legal AI terms, in plain English.

The vocabulary you’ll meet when evaluating AI for your practice - defined without the marketing.

AWS Bedrock

Amazon Web Services' managed service for accessing foundation AI models within the AWS infrastructure boundary, enabling organisations to use AI without transmitting data to external model providers.

Agentic AI

An AI system capable of taking a sequence of actions toward a goal — making decisions, using tools, and executing multi-step processes — with limited or no human intervention at each step.

Human-in-the-loop

An AI design principle where a human reviews, approves, or overrides AI decisions before they take effect, preserving professional responsibility for consequential outcomes.

LLM Hallucination

When an AI language model generates information that is plausible-sounding but factually incorrect, including fabricated case citations, invented statutes, or non-existent legal authority.

Private LLM

A large language model deployed within a controlled, organisation-specific environment where data does not leave the organisation's infrastructure and is not used to train shared models.

A security vulnerability in AI systems where malicious instructions embedded in user-provided content override the system's intended instructions, causing the AI to behave in unintended or harmful ways.

RAG (Retrieval-Augmented Generation)

An AI architecture that retrieves relevant documents from a knowledge base before generating a response, grounding outputs in specific source material rather than relying solely on training data.

Zero Data Retention (ZDR)

A contractual and technical arrangement with an AI provider whereby data submitted to the model is not stored, logged, or used for training after the inference request is complete.

Ready to level the playing field?

See LegalScout on your own contracts in a 20-minute live walkthrough. No pressure. No procurement deck.

Get started

Australian owned · Hosted in AWS Sydney · 24/7 support