Is AI making creative work more similar?
Some studies find that people using generative AI produce more similar work, even when individual outputs improve.
A 2026 meta-analysis of 19 studies found a small but statistically significant homogenising effect in human and AI co-creation. The authors distinguish reduced diversity across outputs from a simple loss of individual creativity. Tilburg University’s research record
The studies cover particular tasks and tools. They do not establish that every brand using AI becomes less distinctive. Where the underlying experiments are more than three months old, rapid model development also limits their value as a measure of current capability.
What makes AI-generated marketing generic?
Generic marketing can result when the input describes familiar category conventions without explaining the brand’s particular offer, audience or character.
A model can produce fluent copy or polished images from that material while leaving little that identifies one business. Reusing the same references and accepting similar initial outputs can further narrow the work considered.
These are properties of the brief and production process, not a verdict on the tool alone. AI can also help explore alternatives when the source material contains useful differences and the review examines whether those differences survive.
How can AI support distinctive brand work?
AI can support distinctive work by developing ideas around the brand’s own evidence, assets and perspective, with people assessing the result.
The relevant inputs may include customer observations, a specific product advantage, established visual assets or a recognisable way of speaking. Their usefulness depends on whether they are true, meaningful to the audience and appropriate to the assignment.
More variations do not necessarily mean more distinct ideas. Different layouts or wording may express the same underlying thought. Comparing the meaning and role of each route reveals more than counting the outputs.
How can a business assess whether its AI work is becoming generic?
It can examine whether the finished work remains identifiable, relevant and meaningfully different from alternatives in the category.
Brand recognition, accurate use of distinctive assets and audience response answer different parts of that question. Internal preference alone cannot establish what customers notice or remember.
A comparison across campaigns can reveal repeated conventions, while testing finished assets can examine response in a particular context. Neither produces a universal score for distinctiveness. The interpretation depends on the brand’s objectives and the evidence collected.