AI terms, clearly explained
What does a term mean, and why does it matter for your organisation? For each term, a short definition, the explanation and what it means in practice.
Each term starts with a definition that stands on its own. Below it, you’ll find what it comes down to in practice, and where to read more.
Agentic AI
AI that plans and carries out steps on its own to reach a goal, using tools and other systems.
Read the definitionAI Act
The European law that classifies AI by risk, with rules for those who build AI and those who use it.
Read the definitionAI agent
An AI system that takes steps on its own to reach a goal, rather than just answering.
Read the definitionAI governance
The roles, rules and controls an organisation uses to decide how it uses AI.
Read the definitionAI hallucination
Output from an AI system that sounds convincing, but is wrong or made up.
Read the definitionAI literacy
Knowing what AI can and can’t do and being able to judge its output. The AI Act asks employers to support it.
Read the definitionAI Opportunity Audit
Upfront research that establishes, per process, which AI applications are feasible and what comes first.
Read the definitionAI readiness
How prepared your organisation is to put AI to work: data, processes, people and governance.
Read the definitionAI roadmap
A phased plan: which AI application when, who owns it and when you decide.
Read the definitionAPI
The fixed entry point a software package opens up, so other software can read or change its data.
Read the definitionDeployer
The AI Act role of an organisation that uses an AI system built by someone else.
Read the definitionGenerative AI
AI that creates new content, such as text, images or code, from an instruction in plain language.
Read the definitionHigh-risk AI system
AI that helps make decisions about people, with the strictest requirements of the AI Act.
Read the definitionLarge language model (LLM)
The language model behind ChatGPT, Claude and Copilot: it predicts which text logically follows.
Read the definitionMachine learning
AI that learns patterns from examples, instead of following rules someone has written down.
Read the definitionModel Context Protocol (MCP)
An open standard that lets an AI assistant fetch data and take actions in other software.
Read the definitionProcess mining
Using the logs in your systems to reconstruct how a process really runs.
Read the definitionPrompt injection
An attack using instructions hidden in data that an AI system processes.
Read the definitionRAG (retrieval-augmented generation)
An AI assistant that searches your own documents first and then answers, with the source attached.
Read the definitionRPA vs AI
Fixed rules that always do the same thing, versus AI that interprets but can vary.
Read the definitionSaaS (software as a service)
Software you use over the internet and rent, instead of installing and managing it yourself.
Read the definition