AI Glossary

A software system that uses AI to autonomously perform tasks or make decisions on behalf of a user, often by combining reasoning with the ability to take actions.

Systematic errors in an AI system's outputs that unfairly favor or disadvantage particular groups, often originating from patterns in the data used to train it.

The study and practice of ensuring AI systems are developed and used fairly, transparently, and responsibly, minimizing bias and harm.

The policies, processes, and oversight structures organizations put in place to ensure AI systems are used responsibly, safely, and in line with legal and ethical standards.

A defined set of steps or rules a computer follows to solve a problem or complete a task — the underlying logic behind any AI system.

A defined way for different software systems to communicate with each other; most AI tools are accessed by applications through an API.

Computer systems designed to perform tasks that typically require human intelligence, such as understanding language, recognizing patterns, or making decisions.

The use of technology to perform tasks with minimal human intervention; AI expands what can be automated by handling tasks that require judgment or pattern recognition, not just fixed rules.

The delivery of computing resources — like processing power and storage — over the internet, which most modern AI systems rely on to train and run at scale.

A field of AI focused on enabling computers to interpret and understand visual information from images and video.

The amount of text an AI model can consider at once when generating a response, measured in tokens.

The process of examining data sets to draw conclusions and support decision-making, often a foundation for applying AI effectively in business.

The set of practices and policies that ensure data is accurate, secure, and used appropriately across an organization — a foundation for reliable AI and analytics.

A subset of machine learning that uses neural networks with many layers to learn increasingly abstract representations of data, enabling tasks like image recognition and language understanding.

The degree to which an AI system's decisions or outputs can be understood and explained in terms a person can follow — an important factor in trust and accountability.

The process of further training an existing AI model on a smaller, specialized dataset to adapt it for a specific task or domain.

AI systems that create new content — text, images, audio, or code — rather than only analyzing or classifying existing data.

Example: A tool that writes a marketing email from a short description is using generative AI.

An instance where a generative AI model produces information that sounds plausible but is factually incorrect or fabricated.

A neural network trained on vast amounts of text, capable of understanding and generating human language — the technology behind tools like ChatGPT.

A branch of AI where systems learn patterns from data rather than following explicitly programmed rules, improving their performance as they see more examples.

The trained system that results from a machine learning process — the output that actually makes predictions or generates content once training is complete.

A field of AI focused on enabling computers to understand, interpret, and generate human language.

A machine learning model loosely inspired by the structure of the brain, made up of layers of connected nodes that learn to recognize patterns in data.

The use of historical data, statistics, and machine learning to forecast future outcomes or trends.

The practice of crafting inputs to an AI system — especially an LLM — to reliably get useful, accurate, and relevant outputs.

A machine learning approach where a system learns by trial and error, receiving rewards or penalties based on the actions it takes.

A technique that improves an AI model's responses by first retrieving relevant information from an external source, then using it to generate a more accurate, grounded answer.

A natural language processing technique that determines the emotional tone — positive, negative, or neutral — expressed in a piece of text.

A machine learning approach where a model learns from labeled examples — data paired with the correct answer — to make predictions on new, unseen data.

A unit of text — often a word or part of a word — that a language model processes; the basic building block LLMs use to read and generate language.

The dataset used to teach a machine learning model to recognize patterns and make predictions; the quality and scope of training data directly shapes what a model can do.

A machine learning approach where a model finds patterns or groupings in data without labeled examples to learn from.