AI glossary
Generative AI
AI that creates new content — text, images, audio, video or code — instead of only classifying or predicting from existing data.
Generative AI learns the patterns of its training data well enough to produce new examples that follow them. Language models generate text and code; diffusion models and similar systems generate images and video; others generate speech and music.
The outputs are new, but they can reproduce biases or errors from the training data, and their use raises questions about copyright and disclosure.
Example: From one idea, “introduce a neighbourhood café”, generative AI can write the website copy, create a logo, generate a photo of the place and read the announcement aloud.
In practice
- Use it as a draft: check facts, tone and rights before publishing.
- Say when content is AI-made if it could mislead; in the EU it is mandatory in some cases.
- Do not put confidential information into tools that do not guarantee how they handle it.
How a model generates text, explained: how LLMs work.
