Reference

AI Glossary

Essential artificial intelligence terms explained clearly, in English. For professionals, managers and teams who want to understand the technology they already use every day.

A B C D E F G H I J K L M N O P Q R S T U V W X Y Z

A

AI Artificial Intelligence Artificial Intelligence

A field of computer science dedicated to creating systems capable of performing tasks that normally require human intelligence, such as language understanding, pattern recognition and decision-making.

AI Agent Artificial Intelligence

An autonomous program that executes tasks, makes decisions and interacts with tools on behalf of a user, based on defined objectives.

Example: An AI agent can monitor an email inbox, reply to routine messages and schedule meetings without human intervention.
AI Literacy Regulation

The ability to understand how AI systems work, their risks and limitations, and how to use them responsibly. Required by Article 4 of the EU AI Act.

Legal context: Since 2 February 2025, companies that use AI systems must ensure that their staff have a sufficient level of AI literacy.
API Application Programming Interface Technology

An interface that enables communication between different software systems. It defines rules and formats so that applications can exchange data in a structured way.

Example: When a weather app shows updated data, it is using an API to fetch that information from an external server.
Automation Operations

The execution of tasks or processes without direct human intervention, through rules or artificial intelligence.

Example: Automatically sending a confirmation email after a booking is the simplest form of automation.

C

Chatbot Artificial Intelligence

A program that simulates conversations with users through text or voice, using predefined rules or language models.

Example: A bank's virtual assistant that answers questions about balance and transfers is a chatbot.
Cloud Computing Technology

A computing model that provides access to resources (servers, storage, applications) via the internet, without local infrastructure.

Example: Using Google Drive to store files instead of keeping them on a local hard drive is a form of cloud computing.

D

Deep Learning Machine Learning

A subfield of machine learning that uses neural networks with multiple layers to learn complex patterns in large volumes of data.

Example: Facial recognition on a smartphone uses deep learning to identify faces in photographs.

E

EU AI Act European Union Artificial Intelligence Act Regulation

Regulation (EU) 2024/1689 of the European Parliament, the first comprehensive legal framework for artificial intelligence. It partially entered into force on 2 February 2025.

Note: Enforcement of Article 4 (AI literacy) begins on 2 August 2026. Learn more.

F

Fine-tuning Machine Learning

The process of adapting a pre-trained AI model to a specific task or domain, using additional, more focused data.

Example: Taking a general language model and training it with medical data so that it responds better to health-related questions.

G

GPT Generative Pre-trained Transformer AI Models

A language model architecture developed by OpenAI, trained to generate text. The foundation of ChatGPT.

L

Language Model AI Models

An AI system trained to understand and generate text in natural language.

LLM Large Language Model AI Models

A large-scale language model, trained on large volumes of text, capable of generating, summarising and understanding natural language. Examples: GPT-4, Claude, Gemini, LLaMA.

M

Machine Learning Artificial Intelligence

A subfield of artificial intelligence in which systems learn from data, identifying patterns and improving performance without being explicitly programmed for each task.

Example: A spam filter that learns to distinguish legitimate emails from junk based on the emails a user marks as spam.

N

NLP Natural Language Processing Artificial Intelligence

An area of AI dedicated to the interaction between computers and human language: understanding, generation and translation of text.

O

Open Source Technology

Software whose source code is public and can be freely used, modified and distributed, within the terms of the licence.

Example: Linux is the best-known example of open source software. In AI, models such as LLaMA (Meta) are made available as open source.

P

Prompt Artificial Intelligence

An instruction or input text given to an AI model to obtain a specific response or result.

Example: Typing "Summarise this document in 3 points" in ChatGPT is giving a prompt to the model.

R

RAG Retrieval-Augmented Generation Artificial Intelligence

A technique that combines information retrieval from documents with text generation by an LLM, enabling responses based on specific, up-to-date data.

Example: An AI assistant that consults a company's internal manuals before answering an employee's question is using RAG.
Risk Level (EU AI Act) Regulation

Classification of AI systems into four categories: unacceptable risk (prohibited), high, limited and minimal. It defines the applicable regulatory obligations.

The four categories: Unacceptable risk (e.g. social scoring) — prohibited. High (e.g. AI-based recruitment) — strict requirements. Limited (e.g. chatbots) — transparency obligations. Minimal (e.g. spam filters) — no additional restrictions.

S

SaaS Software as a Service Technology

A software distribution model where the application is accessed via the internet, with subscription-based payment, without local installation.

Example: Gmail, Slack and Notion are examples of SaaS — accessed via a browser, paid monthly, with nothing to install.

T

Token AI Models

The basic unit of text processed by a language model. It can be a word, part of a word or a punctuation character. Models have token limits per request.

Example: The sentence "Hello, world!" can be split into 4 tokens: "Hello", ",", "world", "!". Models charge based on the number of tokens processed.

V

Vertical Business

A specific sector or industry (e.g. healthcare, real estate, fitness). At D'One, agents are configured per vertical to understand the particularities of each sector.

Example: A dental clinic and a real estate agency are different verticals — each with its own processes, language and needs.

Want to know how AI applies to your business?

D'One sets up autonomous AI agents for SMEs — scheduling, invoicing, content and more. Fully managed, fully compliant with the EU AI Act.

Get in touch