AICreative AIStory 02
Image prompts can acquire stereotypes invisibly
What changedA study of image-generation products finds that hidden prompt revision can introduce cultural stereotypes before the image model runs. The result shifts part of the audit target upstream: a polished final picture may reflect changes the user never saw. The tested products, languages and prompts are finite, so the mechanism should not be generalized to every generator.
The useful part
Why it matters
Product audits need to preserve and inspect the final rewritten prompt, because evaluating only the visible request and output can miss where bias entered.
Worth doing
What to do next
Log prompt transformations, disclose material rewrites and include multilingual paired tests in image-generation evaluations.
Keep in mind
Good to know
The study covers a bounded set of systems, languages and prompts rather than every image generator or deployment configuration.
Evidence