« Back to Glossary Index

AI image generation is the creation of pictures from a written description, known as a prompt. Type something like “a cosy coffee shop at sunset, photorealistic,” and the model produces an original image built to match — no camera, illustrator, or stock library required.

Most modern image generators rely on a technique called diffusion. The model starts with pure random noise and gradually refines it, step by step, into a coherent picture matching the prompt. It gets there by having learned from millions of paired images and captions, so it understands how language maps onto visual detail.

  • Midjourney — striking, artistic results, run through Discord
  • DALL-E 3 — built into ChatGPT Plus, strong at literal prompt interpretation
  • Stable Diffusion — open source, highly customisable, can even run locally
  • Adobe Firefly — trained on licensed imagery, considered safer for commercial use
  • Ideogram — notably good at rendering legible text inside images
  • Custom hero images and blog illustrations
  • Product mockups and lifestyle shots
  • Icons and other graphic elements
  • Social media visuals

The legal picture around AI-generated imagery is still settling. For anything commercial, it's worth favouring tools trained on licensed data — Adobe Firefly, for instance — or reading each provider's terms of service closely before publishing.

« Back to Index
Share This