AI-assisted food design: co-creation with young consumers

Advancements in food design and technology are reshaping creative product development by driving highly collaborative workflows. Generative Artificial Intelligence (AI) addresses a critical challenge in traditional consumer research: the barrier to verbalizing subjective sensory ideas. By rendering abstract user insights into explicit visual representations, this technology bridges the gap between consumer imagination and initial product execution.
This study evaluated a novel food design framework that embeds generative AI as a visual mediator during the participatory co-creation of an innovative alfajor concept tailored to young consumers.
A multi-stage participatory design approach was deployed. First, an online ideation survey with frequent consumers (n=124; 61% women, average age 28.0±3.4 years) explored novel configurations of the alfajor’s core components. From these inputs, nine high-potential innovative concepts were defined by calculating an Innovation Index based on market novelty. Second, textual concepts were translated into realistic visual stimuli using generative AI (GPT-5) with standardized prompts. Finally, a second group of consumers (n=343; 75.3% women, average age 24.6±4.1 years) evaluated these visual prototypes to measure expected sensory acceptability (9-point hedonic scale), purchase intention (9-point scale), willingness to pay (categorical scale), and text-image conceptual congruence (7-point scale).
The findings validate generative AI as a novel and efficient asset for agile food design. Participants reported an 80% image-concept congruence (5.2±1.6 out of 7). Purchase intent and congruence demonstrated a positive, weak correlation (r = 0.23). Expected acceptability ranged from a maximum of 7.0±1.9 to a minimum of 6.0±2.6, both being comparable to the value of 6 standardly used in sensory evaluation to indicate an acceptable product. While no physical prototypes were used, meaning the results reflect expected sensory acceptance, this visual mediation successfully transformed textual descriptions into anticipated consumption experiences. Ultimately, AI-assisted co-creation significantly streamlines early-stage food product design by delivering promising, user-validated visual concepts.