Shoppers rate products lower when AI designed them; a framing fix halves the penalty

The penalty grew when AI involvement was visible in the final product, was larger for expressive products than functional ones, and was larger among consumers primed to see themselves as independent. The authors link it partly to a new measure, perceived authenticity of creative intention: whether people infer a human pursued a genuine creative vision.

A disclosure approach called creative-intent framing narrowed the penalty by roughly half, also through the authenticity mechanism. The authors present it as a scalable disclosure design principle for marketers, though the paper reports experimental consumer responses rather than real sales outcomes.

The study tests a 2-by-2 taxonomy it calls algorithmic provenance. One axis is whether AI acted as an agent (the designer) or an assistant (a tool for a human designer); the other is whether AI involvement is visible or invisible in the finished product. Identical products were evaluated differently depending on their place in that grid.

The main finding is an agentic penalty: when AI is the designer, consumers report lower product attitudes, quality ratings, purchase intentions and willingness to pay than when it assists a human. Visibility of AI in the final product made the penalty larger.

Authors identify perceived authenticity of creative intention as a partial mediator, capturing whether consumers infer a human creator pursued a genuine creative vision. The penalty was larger for expressive than functional products and among consumers primed with independent self-construal.

Creative-intent framing, a disclosure approach marketers could scale, reduced the penalty by about half through the same authenticity mechanism. The authors say the results refine theories of algorithmic aversion and give marketers a disclosure design principle.