AI Learn
AI Lab
Glossary
Dark Mode
Sign in
Ask Forge
AI Lab
Glossary
Glossary
AI jargon, translated.
Agent
An AI system that can plan steps and use tools to reach a goal, rather than just answering one message.
Lesson: Agents and autonomy →
AI (Artificial Intelligence)
Software that performs tasks that normally need human intelligence, such as understanding language or spotting patterns.
Lesson: What is AI and how chat models work →
AI champion
Our team
A colleague in each team who helps others try AI safely and shares good use cases.
Lesson: How we use AI here →
API
A way for one piece of software to talk to another in a structured way.
Lesson: MCPs: connecting AI to your tools →
Approved AI tool
Our team
An AI tool the company has checked for security and data protection and allows staff to use with work information.
Lesson: How we use AI here →
Automation
A workflow that runs by itself when something happens (a trigger), often with AI handling the judgement steps.
Lesson: Automations: AI inside your processes →
Autonomy
How much an AI is allowed to do on its own before a human checks or approves.
Lesson: Agents and autonomy →
Context window
How much text an AI model can 'see' at once, including your conversation and any documents.
Lesson: What is AI and how chat models work →
Copilot
Our team
Microsoft's AI assistant built into Outlook, Word, Excel, Teams and other Office apps.
Lesson: Use case: meeting notes and actions →
Data minimisation
A GDPR principle: only use the personal data you actually need for the task.
Lesson: Data compliance and AI →
Deepfake
AI-generated audio, image or video that imitates a real person.
Lesson: Data and cyber security with AI →
Embedding
A list of numbers that captures the meaning of text, used to find similar content.
Lesson: More things AI can do →
Fine-tuning
Further training a model on specific examples to specialise it.
Lesson: What is AI and how chat models work →
GDPR
UK/EU data protection law governing how personal data is collected, used and stored.
Lesson: Data compliance and AI →
Guardrails
Rules and checks that keep an AI system's behaviour safe and on-task.
Lesson: Agents and autonomy →
Hallucination
When an AI confidently states something false or made up.
Lesson: What is AI and how chat models work →
Human in the loop
A design where a person reviews or approves AI actions at key points.
Lesson: Automations: AI inside your processes →
LLM (Large Language Model)
An AI model trained on huge amounts of text to predict and generate language. ChatGPT, Claude and Gemini are built on LLMs.
Lesson: What is AI and how chat models work →
MCP (Model Context Protocol)
An open standard that lets AI assistants plug into tools and data sources (calendar, CRM, files) in a consistent way — like USB-C for AI.
Lesson: MCPs: connecting AI to your tools →
Multimodal
An AI that works with more than text — images, audio, video.
Lesson: More things AI can do →
Personal data
Any information that identifies a living person, such as name, email, or employee ID.
Lesson: Data compliance and AI →
Phishing
Fraudulent messages that trick people into revealing information or clicking malicious links; AI makes them more convincing.
Lesson: Data and cyber security with AI →
Power Automate
Our team
Microsoft's tool for building automations between apps — often used to run AI steps in our workflows.
Lesson: Use case: sorting the shared inbox →
Process automation
Our team
Using software (often with an AI step) to run a repetitive, rule-based business process end to end, with people handling the exceptions.
Lesson: Use case: process automation →
Prompt
The instruction or question you give an AI.
Lesson: Using chat well: prompting basics →
Prompt injection
Hidden instructions inside content (an email, webpage) that try to hijack an AI agent.
Lesson: Data and cyber security with AI →
Prompt library
Our team
A shared collection of tested prompts for common team tasks, so nobody starts from scratch.
Lesson: Use case: meeting notes and actions →
RAG (Retrieval-Augmented Generation)
Giving an AI relevant documents at question time so it answers from your sources.
Lesson: More things AI can do →
System prompt
Background instructions that set an AI's role and rules before the user speaks.
Lesson: Using chat well: prompting basics →
Token
A chunk of text (roughly ¾ of a word) that AI models read and write; usage is often priced per token.
Lesson: What is AI and how chat models work →
Tool use
When an AI calls an external function — search, calculator, send email — to get something done.
Lesson: Agents and autonomy →
Use case
Our team
A specific task where AI adds value, described with who does it, how often, and what good looks like.
Lesson: How we use AI here →
Vibe coding
Building software by describing what you want in plain language and letting AI write the code.
Lesson: Vibe coding: building by describing →
Zero data retention
A setting where the AI provider does not store your prompts or outputs.
Lesson: Data compliance and AI →
