While research on AI in management has dedicated significant attention to AI-using producers, little work has examined audience members’ evaluation of AI-generated products in markets. To fill this gap, we develop a theory of audience members’ evaluation of AI-generated products, building on work on authenticity. We argue that AI-generated products have two key characteristics associated with their specific production process which categorically comparable products do not have. First, they are inscrutable, and hence suffer from producer ambiguity. Second, they are embedded in a cultural antinomy that contrasts human enhancement with human replacement by AI, and hence suffer from moral ambiguity. The two ambiguities result in an authenticity discount for AI-generated products relative to non-AI-generated products belonging to the same product category and of comparable quality. The use of generative AI in the production of peripheral rather than core features reduces producer ambiguity while the use of human supervision of generative AI in the production process rather than of an autonomous AI reduces moral ambiguity. Both strategies may even result in an authenticity premium compared to categorically comparable products as they help AI-generated products to appear not only as authentic but also as reflecting creative self-expression.
