Image generation is artificial intelligence that creates images from a text description – a prompt. Tools such as Midjourney, DALL·E and the built-in features of the office suites have made visual production available to everyone.
The opportunity is obvious: illustrations, moodboards and variations without waiting and without a budget. But precisely because everyone can, the visual expression becomes a strategic choice. A brand that generates in every direction loses its face faster than it realises.
Then there are the rights: who owns a generated image, what was it trained on, and may it be used commercially? The answers depend on the tool and the agreements – and should be settled before the images reach the channels.
In practice
The organisations that succeed define an image style before they open up the tools: motifs, mood, what AI may be used for – and what must always be genuinely photographed. That way, generation becomes a reinforcement of the identity instead of a dilution.
How to explain it to management
»Image generation makes it cheap to produce images – and expensive not to care what they look like.«
You can read how it fits with the rest of the creative work on the page about AI in creative work.